When and with Whom to Ally? the Influence of Cm Strategic Alliances

ASSOCIATION FOR CONSUMER RESEARCH

Labovitz School of Business & Economics, University of Minnesota Duluth, 11 E. Superior Street, Suite 210, Duluth, MN 55802

When and With Whom to Ally? the Influence of Cm Strategic Alliances on Consumer Purchase Behavior Chien-Wei (Wilson) Lin, State University of New York at Oneonta, USA Qi Wang, State University of New York at Binghamton, USA

Five studies investigate the impact of two types of strategic alliances in cause-related marketing: within- and cross-industry alliances. We find consumers perceive within- (vs. cross-) industry alliance as a high entitativity group with strong commitment to a cause, resulting in greater purchase intentions. Familiarity and cause involvement moderate the effect.

[to cite]: Chien-Wei (Wilson) Lin and Qi Wang (2017) ,"When and With Whom to Ally? the Influence of Cm Strategic Alliances on Consumer Purchase Behavior", in NA - Advances in Consumer Research Volume 45, eds. Ayelet Gneezy, Vladas Griskevicius, and Patti Williams, Duluth, MN : Association for Consumer Research, Pages: 749-450.

[url]: http://www.acrwebsite.org/volumes/1024126/volumes/v45/NA-45

[copyright notice]: This work is copyrighted by The Association for Consumer Research. For permission to copy or use this work in whole or in part, please contact the Copyright Clearance Center at http://www.copyright.com/. When and With Whom to Ally? The Influence of CM Strategic Alliances on Consumer Purchase Behavior Chien-Wei (Wilson) Lin, State University of New York at Oneonta, USA Qi Wang, State University of New York at Binghamton, USA

EXTENDED ABSTRACT Therefore, within-industry alliance message can be more desirable According to the IEG Sponsorship Report (2016), U.S. cause- for low- (vs. high-) cause involvement individuals. related marketing (CM) spending reached $1.92 billion in 2015, Results from four studies and an empirical data support our hy- representing a 135 percent increase from 2002. Recently, it is in- potheses. Study 1 employs a 3 (alliance: within, cross, no) x 2 (famil- creasingly common to find firms involved the strategic alliances with iarity: well-known, unknown) between-subjects design. Participants others firms to execute the CM program—CM alliance (Varadarajan were first asked to examine a recent company announcement, named and Menon 1988). For example, General Mills teamed up with the “The Fighting Hunger Project”: During the next three months the other food companies (e.g., ConAgra Foods, Kraft Foods, and Kel- firm(s) will donate 5% of the proceeds from the sales of the partici- logg) to donate meals to Feeding America. Likewise, the Product pating products to help fight hunger. We chose General Mills as the (RED) campaign engaged companies such as Apple, Gap, and Bel- well-known target company and a fictitious food company named vedere Vodka to contribute 50% of profits from designated products popp as the unknown target company. In the within-industry alli- to provide medicine to AIDS patients in Africa. ance condition, the company worked together with the other three We refer to the former type of CM alliance as within-industry companies from the same food industry (e.g., Kraft foods, Kellogg’s, alliance (General Milles), the latter as cross-industry alliance (RED). and Nestle) and in the cross-industry alliance condition, the company While solo CM campaign remains dominant in the marketplace (138 worked together with the other companies across different industries campaigns, 41 percent), both within- (83 campaigns, 25 percent) and (e.g., Ford, Motorola, and Gap). In the no alliance condition, the cross- (113 campaigns, 34 percent) industry alliances are viewed as company worked exclusively with the cause. As expected, partici- important strategic alternatives for many firms (Engage for Good pants were more likely to buy the products for within- than cross- 2015). The above raises important questions that have not yet been industry alliance when the company is unknown, but not when the fully investigated: With whom should firms ally to make CM more company is well-known. effective? Furthermore, the impact of within- and cross-industry al- Study 2 tested the moderating effect of cause involvement on liances may depend upon firm and consumer characteristics. When consumers’ preference toward cross-industry alliance products. The firms should ally to make CM more effective? between-subjects design and stimuli were the same as before. Par- A CM alliance can play a role in how a firm’s commitment to ticipants indicated their cause involvement in the end (Grau and a cause is perceived, which can affect acceptance of the good deed Folse 2007). Results replicate the finding in study 1 and this effect is (Van den Brink et al. 2004). According to attribution theory (Heider diminished for high cause-involvement individuals, supporting the 1958), when consumers perceive that a company is making a signifi- moderating role of cause involvement. cant effort to a cause, they are more likely to infer altruistic motives Study 3 ruled out company-cause fit as an alternative account on the part of the sponsor, and thoughts of corporate profits associ- (i.e., lower fit between fighting hunger and partners from different ated with the sponsorship act can be minimized (Rifon et al. 2004). industries than from the same food industry). Study 3 employed We predict that a within-industry alliance may elicit perceptions of a 2 (alliance: within, cross) x 2 (fit: high, low) x 3 (product: food, high entitativity group with strong commitment to the cause and detergent, calculators) mixed design with alliance and fit as the be- thus increase customer preferences to a firm’s products. Entitativity tween-subjects factors and product as the within-subjects factor. We has been defined as “the perception that an aggregate of individuals generated the high- and low-fit conditions for each product category is bonded together in some way to constitute a group” (Hamilton, (confirmed by a pretest). As expected, within- (vs. cross-) industry Sherman, and Castelli 2002, p. 141). High entitativity groups share alliance leads to greater purchase intentions, regardless of high (or properties such as similarity, common movement, and common goals low) cause fit with the company. (Campbell 1958; Lickel et al. 2000) and their members are perceived Study 4 examined the underlying mechanism in a 2 (alliance) as active participants in carrying out plans to achieve collective ob- x 2 (product) between-subjects design. We measured participants’ jectives (Brewer et al. 2004). These characteristics are related to the willingness-to-pay (WTP) and purchase intentions. To test the un- factors of greater perceived commitment: the amount of input, the derlying mechanism, participants responded to items including the durability of the association, and the consistency (stability) of input perceptions of corporate commitment to the cause and entitativity (Dwyer, Schurr, and Oh 1987). regarding the alliance. As expected, within- (vs. cross-) industry al- The present work further examines firm and consumer -char liance leads to stronger perceptions of entitativity, which leads to acteristics that moderate the CM alliance effect. For a well-known greater perceptions of corporate commitment to the cause. The great- company, the degree of liking for the brand is established (Bettman er perceived corporate commitment in turn leads to a greater WTP. and Sujan 1987) so forming a CM alliance will not benefit brand Having observed the commitment issue of cross-industry alliance, attitude much. However, for an unknown company, consumers will we verified it using an empirical data. Using data from an online CM heavily scrutinize its motive to make a judgment due to a lack of trust news website, we found that firms are less likely to continue cross- (Ellen et al. 2006). Thus, we predict that a within-industry alliance (vs. within-) industry alliance. All effects were significant atp < .05. can enhance consumers’ confidence in the unknown firm’s intrinsic In sum, this research demonstrates novel effects of using CM motive, thereby increasing CM product preference. Further, Grau alliance by showing when and what types of CM alliance are effec- and Folse (2007) find that compared with high-cause involvement in- tive in influencing consumer response. By showing that entitativity dividuals, low-cause involvement individuals rely more on CM mes- affects perceived cause commitment, we demonstrate a novel process sage cues to produce favorable attitudes and participation intentions. through which emotional and charitable responses can be increased.

Advances in Consumer Research 749 Volume 45, ©2017 750 / When and With Whom to Ally? The Influence of CM Strategic Alliances on Consumer Purchase Behavior REFERENCES Hamilton, David L., Steven J. Sherman, and Luigi Castelli (2002), Bettman, James R., and Mita Sujan (1987), “Effects of Framing on “A Group by any Other Name—The Role of Entitativity in Evaluation of Comparable and Noncomparable Alternatives Group Perception,” European Review of Social Psychology, by Expert and Novice Consumers,” Journal of Consumer 12 (April), 139-166. Research, 14 (September), 141-154. Heider, Fritz (1958), The Psychology of Interpersonal Relations, Brewer, Marilynn B., Ying-Yi Hong, and Qiong Li (2004), New York: Wiley. Dynamic Entitativity: Perceiving Groups as Actors. In V. International Event Group (IEG) (2016), Sponsorship spending in Yzerbyt, C. M. Judd, & O. Corneille (Eds.), The psychology North America, Chicago: IEG Sponsorship Report. of group perception: Perceived variability, entitativity, and Lickel, Brian, David L. Hamilton, Grazyna Wieczorkowska, Amy essentialism (pp. 25–38). New York: Psychology Press. Lewis, Steven J. Sherman, and A. Neville Uhles (2000), Campbell, Donald T. (1958), “Common Fate, Similarity, and Other “Varieties of Groups and the Perception of Group Entitativity,” Indices of The Status of Aggregates of Persons as Social Journal of Personality and Social Psychology, 78 (February), Entities,” Behavioral Science, 3 (1), 14-25. 223–46. Dwyer, F. Robert, Paul H. Schurr, and Sejo Oh (1987), “Developing Rifon, Nora J., Sejung Marina Choi, Carrie S. Trimble, and Hairong Buyer-Seller Relationships,” Journal of Marketing, 51 (April), Li (2004), “Congruence Effects in Sponsorship: The Mediating 11-27. Role of Sponsor Credibility and Consumer Attributions of Ellen, Pam Scholder, Deborah J. Webb, and Lois A. Mohr (2006), Sponsor Motive,” Journal of Advertising, 33 (March), 30-42. “Building Corporate Associations: Consumer Attributions for Varadarajan, P. Rajan, and Anil Menon (1988), “Cause-Related Corporate Socially Responsible Programs,” Journal of the Marketing: A Coalignment of Marketing Strategy and Academy of Marketing Science, 34 (April), 147-157. Corporate Philanthropy,” Journal of Marketing, 53 (July), Engage for Good (2015), News and Articles, Retrieved March 08, 58-74. 2016, from http://engageforgood.com/resources/articles/. Van den Brink, Douwe, Gaby Odekerken-Schröder, and Pieter Grau, Stacy Landreth, and Judith Anne Garretson Folse (2007), Pauwels (2006), “The Effect of Strategic and Tactical Cause- “Cause-Related Marketing (CRM): The Influence of Donation Related Marketing on Consumers’ Brand Loyalty,” Journal of Proximity and Message-Framing Cues on the Less-Involved Consumer Marketing, 23 (1), 15-25. Consumer,” Journal of Advertising, 36 (March), 19-33. The Effect of Social Exclusion on Consumer Shoplifting Christopher Ling, University of South Carolina, USA Thomas Kramer, University of California, Riverside, USA

EXTENDED ABSTRACT lifted more pieces of chocolate than did non-shoplifters (M = .21; Consumer shoplifting, the theft of merchandise from retail es- F(1, 80) = 3.99, p < .05). tablishments, is a serious problem for retailers, representing more Study 2 was a 2 (social acceptance: inclusion versus exclusion) than $16 billion worth of lost sales each year (Allen 2014; National x 2 (shoplifting history: shoplifter versus non-shoplifter) between- Learning & Resource Center 2014). Surprisingly, shoplifting is quite subjects design with social acceptance manipulated and with shop- a ubiquitous phenomenon: an astounding 60% of consumers admit lifting history measured. Social acceptance was manipulated using to having shoplifted at some point in their lifetime (Baumer and the Cyberball game as in study 1. Next in an ostensibly unrelated Rosenbaum 1984). Most shoplifters are “non-professionals” without study, all participants were instructed to imagine that they were alone prior criminal record, who steal not because of financial need but in a department store with an opportunity to steal a pair of jeans because of social influences (National Learning & Resource Center worth about $150. 2014). It is these social influences on which we focus in this research A 2 (Social Acceptance: Social Inclusion vs. Social Exclusion) to advance our understanding of shoplifting. That is, we investigate x 2 (Shoplift) ANCOVA on shoplifting intention, controlling for age, under which conditions, shoplifting intentions and actual shoplifting was significant (F(1,69) = 4.41, p < .05). We also found a main effect behavior may ensue from social exclusion – experiencing a lack of of shoplifting history (F(1,69) = 13.03, p < .001), such that those social connection because of being alone, isolated, or rejected (Bau- individuals who had shoplifted before reported significantly higher meister et al. 2005; Twenge et al. 2001). intention to shoplift the pair of jeans (Ms = 2.26 vs. 1.16). More Shoplifting behavior usually begins when social influence has a importantly, planned contrasts revealed that the socially-excluded particularly strong effect on consumers – in adolescence (Cox, Cox, shoplifters expressed greater intentions to shoplift the jeans than did and Moschis 1990; Mangleburg, Doney, and Bristol 2004), and ado- the socially-excluded non-shoplifters (3.08 vs. .99; F(1, 69) = 13.32, lescents frequently cite social influences to justify their shoplifting p = .001). However, shoplifting history did not impact shoplifting behavior (Cox et al. 1990; Forney, Crutsinger, and Forney 2006). intentions among socially-included participants (1.85 vs. 1.30 for Further, because social exclusion engenders negative affect, we ex- shoplifters vs. non-shoplifters; F(1, 69) = 1.45, p > .20). pect that a driving reason for the effect of social exclusion on shop- The objective of study 3 was to test for the moderating role lifting is that socially-excluded consumers are motivated to engage of product type on our previously found effect in study 2. Partici- in behavior to repair and escape their negative mood through the pants first completed the same social acceptance manipulation from theft of merchandise. Lastly, social exclusion is unlikely to have a study 2. Next, we presented participants with the same shoplifting uniform impact on the shoplifting intentions of all consumers alike. scenario previously used which asked them to indicate their inten- Instead, although socially-excluded (vs. included) consumers gener- tion to shoplift; however, we manipulated whether the target option ally are likely to experience greater negative affect, it is only those was a hedonic (6 pack of energy drinks) or a utilitarian (pack of pain socially-excluded consumers who have experienced shoplifting’s relievers) product. After participants had indicated their intention to mood-enhancing properties (i.e., shoplifters - those who have shop- shoplift, we measured the extent to which shoplifting was perceived lifted before) that should seek to repair their mood by shoplifting. as a mood repair mechanism In this research, we argue that one way to cope with negative af- A 3 (social acceptance: exclusion vs. inclusion vs. control) x 2 fect triggered by social exclusion is through shoplifting because of its (shoplifting history: shoplifter vs. non-shoplifter) x 2 (product type: perceived mood repair properties. The theft of merchandise has been hedonic vs. utilitarian) ANCOVA, controlling for age, on shoplifting linked with excitement and experiential properties (Kallis and Vanier intention was significant;F (2, 595) = 5.08, p < .01. planned contrasts 1985; Sarasalo, Bergman, and Toth 1997), with shoplifting providing revealed that socially-excluded shoplifters (vs. non-shoplifters) ex- both the pleasure from consumption and also from the acquisition of pressed greater intentions to shoplift the hedonic energy drinks (2.16 the product. Some shoplifters, for example, have self-reported ex- vs. 1.19; F(1, 595) = 10.94, p = .001). periencing hedonic pleasure from the act of successfully shoplifting Next, we tested for moderated mediation with the mediator of (Fullerton and Punj 1993). When in a negative mood, socially-ex- shoplifting as perceived mood repair using model 12 from the PRO- cluded (vs. included) consumers may be motivated to engage in devi- CESS macro (Hayes 2013). The index of moderated mediation was ant thrill-seeking in an effort to feel better, and shoplifting constitutes significant (95% CI: .17, 2.01) providing evidence for moderated such a means (Fullerton and Punj 1998). mediation (Hayes 2015). Study 1 was a one-factor, two level (shoplifting history: shoplifter versus non-shoplifter) between-subjects design. Participants REFERENCES arrived to the lab and were socially excluded using Cyberball (Wil- Allen, Kath Granni (2014), “Retailers Estimate Shoplifting, liams and Jarvis 2006). Once the session was over, lab assistants Incidents of Fraud Cost $44 Billion in 2014,” National Retail counted the number of chocolates left in the bowl and the amount Federation, 2014–15, https://nrf.com/media/press-releases/ of money in the jar, so that we could examine if any chocolates had retailers-estimate-shoplifting-incidents-of-fraud-cost-44- been taken without payment. billion-2014. Participants who indicated that they had shoplifted before were Baumeister, Roy F., C. N. DeWall, Natalie J. Ciarocco, and Jean M. defined as shoplifters, whereas those who had not previously shop- Twenge (2005), “Social Exclusion Impairs Self-Regulation.,” lifted were defined as non-shoplifters. A single-factor (shoplifting Journal of Personality and Social Psychology, 88(4), 589–604. history: shoplifter vs. non-shoplifter) ANCOVA, controlling for age Baumer, Terry L and Dennis P Rosenbaum (1984), Combating and gender, found that socially-excluded shoplifters (M = .64) shop- Retail Theft: Programs and Strategies, Butterworth Boston.

Cox, Dena, Anthony D. Cox, and George P. Moschis (1990), Kallis, M J. and Dinoo J. Vanier (1985), “Consumer Shoplifting: “When Consumer Behavior Goes Bad: An Investigation of Orientations and Deterrents,” Journal of Criminal Justice, Adolescent Shoplifting,” Journal of Consumer Research, 13(5), 459–73. 17(2), 149–59. Mangleburg, Tamara F., Patricia M. Doney, and Terry Bristol Forney, William S., Christy Crutsinger, and Judith C. Forney (2004), “Shopping with Friends and Teens’ Susceptibility to (2006), “Self-Concepts and Self-Worth as Predictors of Self- Peer Influence,”Journal of Retailing, 80(2), 101–16. Perception of Morality: Implications for Delinquent Risk National Learning & Resource Center (2014), “Shoplifting Behavior Associated With Shoplifting,” Family and Consumer Statistics,” http://www.shopliftingprevention.org/ Sciences Research Journal, 35(1), 24–43. whatnaspoffers/nrc/publiceducstats.htm. Fullerton, Ronald A. and Girish Punj (1993), “Choosing to Sarasalo, Elina, Bo Bergman, and James Toth (1997), “Theft Misbehave: A Structural Model of Aberrant Consumer Behaviour and Its Consequences among Kelptomaniacs Behavior,” in Advances in Consumer Research, 570–74. and Shoplifters - a Comparative Study,” Forensic Science ——— (1998), “The Unintended Consequences of the Culture International, 86, 193–205. of Consumption: An Historical-Theoretical Analysis of Twenge, Jean M., Roy F. Baumeister, Dianne M. Tice, and Tanja S. Consumer Misbehavior,” Consumption Markets & Culture, Stucke (2001), “If You Can’t Join Them, Beat Them: Effects 1(4), 393–423. of Social Exclusion on Aggressive Behavior.,” Journal of Hayes, Andrew F. (2015), “An Index and Test of Linear Moderated Personality and Social Psychology, 81(6), 1058–69. Mediation,” Multivariate Behavioral Research, 50(1), 1–22. Williams, Kipling D. and Blair Jarvis (2006), “Cyberball: A ——— (2013), Introduction to Mediation, Moderation, and Program for Use in Research on Interpersonal Ostracism and Conditional Process Analysis A Regression-Based Approach, Acceptance.,” Behavior Research Methods, 38(1), 174–80. Guilford Press. The Effect of Facial Resemblance on Product Purchase: The Moderating Role of Mental Construal Fan Liu, Adelphi University, USA Xin He, University of Central Florida, USA Ze Wang, University of Central Florida, USA

EXTENDED ABSTRACT focus on each individual face rather than a group of faces (Trope The significance of human facial resemblance has emerged as and Liberman 2003). In summary, we predict that (a) construal level an important driver of both psychology theories (Zebrowitz, Ki- moderates the effect of facial resemblance on product purchase like- kuchi, and Fellous 2010) and consumer research (e.g. Tanner and lihood, such that facial resemblance enhances purchase likelihood Maeng 2012), which include mate choice, kin selection, and social only among consumers with high-level construals; and (b) the inter- interaction (DeBruine 2004; Hinsz 1989; Kraus and Chen 2010; active effect of facial resemblance and construal level is mediated by Platek et al. 2003, 2004). Although it is well established that people perceived group entitativity. draw inferences about others based on the way they look (e.g. facial morphology), the current literature has mostly focused on perception Methodology of, and behavior towards, individual faces that resemble oneself or The hypotheses were tested in a series of three experiments. one’s familiar others. There is a dearth of research that explores the To prepare for the experimental stimuli, we followed digital morph- inferences derived from the faces of groups (of individuals) presenting to manipulate facial resemblance, also adopted by recent face ed together. In this paper, groups formed by pairs of individual faces research (e.g. Verosky and Todorov 2010). This technique digitally were shown to resemble each other. In particular, we aim to address combines the facial photographs of two (or more) different individu- three research questions: does collective facial appearance of a group als to produce a composite face that represents a weighted average matter? Will facial resemblance among team members bias consum- of the features of all the input faces. By controlling how much each ers’ group-level perception and subsequent decision-making? Which input face contributes to the morph output (anywhere from 0% to type of consumers may be susceptible to face-based biases in judging 100% of the total contribution) we are able to precisely (and objec- groups? tively) vary the degree of facial resemblance among team members in this study’s stimuli samples. In such case, two pairs of faces were Theoretical Development created as experimental stimuli as either the low resemblance group Adding to the literature on face-based perception, this paper ex- (10% morphing) or the high resemblance group (30% morphing). amines the impact of participants’ drawing inference about group fa- In experiments 1, evidence was provided indicating that sub- cial resemblance on participants’ evaluation of the presented groups tle changes in facial resemblance among team members influence along with the product presented by the groups. This relationship de- product purchase likelihood, depending on construal level. One pends on the perceivers’ construal level and springs from inferences hundred and seventy-three undergraduate participants were ran- made based on a group’s faces. As construal level theory illustrates, domly assigned to either the low-resemblance condition or the high- when perceiving several objects within a big picture, perceivers can resemblance condition. Participants were asked to report their intent either have a focus more intently on higher-order goals or on contex- in purchasing the product and complete the Behavior Identification tualized, lower-order details (Liberman, Trope, and Wakslak 2007; Form measuring their chronic construal level (Vallacher and Wegner Trope and Liberman 2003). Similarly, based upon Schwarz and 1989). Among participants with high-level construals, the spotlight Bless’s (1992, 2007) model of exclusion/inclusion as well as Förster analysis revealed a significant and positive effect on purchase likeli- and his coauthors (2008) global local model of social judgment hood (babstract = .68, t = 2.53, p = .01). However, for participants with (GLOMO), high-level construals in a global processing are more low-level construals, there is no significant effect of facial resem- inclusive than in a local processing; such perceivers tend to group blance on purchase likelihood (bconcrete = -.06, t = -.24, ns). the stimulus objects in the same category. Hence, when a high-level The purpose of experiment 2 is to confirm the results using construal is induced, participants process faces globally and have a a different operationalization of construal level. In the high-level tendency to make inferences about group traits based on faces of construal conditions the headline emphasized the end outcomes of the presented group (Liberman, Trope, and Wakslak 2007; Trope and participation. By contrast, in the low-level construal conditions, the Liberman 2003). Accordingly, facial resemblance enhances consum- headline pinpointed the means to achieve the end results. The study ers’ perception of how group members work well with each other. has a 2 (facial resemblance: high versus low) × 2 (construal level: Increasing group facial resemblance leads to greater mental infer- high versus low) between-subjects design. One hundred thirty-three ences about perceived group entitativity (i.e. groupness, homogene- participants from mTurk crowdsourcing services were recruited and ity, cohesiveness). Specifically, facial resemblance at a group level asked to report product purchase likelihood. A two-way ANOVA signals a higher level of intragroup similarity or entitativity, which conducted on purchase likelihood indicated a significant interaction is often associated with increased cooperative intent (Wang, He, and effect (F(1, 130) = 4.26, p < .05). Planned contrasts analyses revealed Liu 2016) or higher likelihood to work toward a common goal (Ip, that consumers at high-level construals tended to buy the product Chiu, and Wan 2006). While there is adequate evidence showing that presented by the high-resemblance versus low-resemblance artists entitative groups benefit from positive team traits that lead to bet- (Mlow = 2.30, Mhigh = 3.08; F(1, 130) = 4.97, p < .05). Among consum- ter team performance (e.g. Hollingshead 1998; Mulvey and Klein ers at low-level contruals, there was no significant difference in pur- 1998), consumers then consistently transfer perceptual group enti- chase likelihood between the facial resemblance conditions (F < 1). tativity, inferred from the evaluation of represented groups of faces, In experiment 3, we operationalized construal level as the de- to the associated presented products (Beckwith and Lehmann 1975; sirability (high-construal) or feasibility (low-construal) of the prod- Wirtz and Bateson 1995). On the contrary, when a low-level con- uct and replicated the findings, In addition, we tested the underlying strual is activated, perceivers tend to conduct local processing and mechanism of perceived group entitativity. Similarly, One hundred

Advances in Consumer Research 753 Volume 45, ©2017 754 / The Effect of Facial Resemblance on Product Purchase: The Moderating Role of Mental Construal and seventy participants were randomly assigned to one of the four Liberman, Nira, Yaacov Trope, and Cheryl Wakslak (2007), experimental conditions. They then reported product purchase in- “Construal Level Theory and Consumer Behavior,” Journal of tent and perceived group entitativity. Consistently, a 2 × 2 ANOVA Consumer Psychology, 17 (2), 113-17. analysis on purchase likelihood was performed to show a significant Mulvey, Paul W., and Howard J. Klein (1998), “The Impact of interaction effect (F(1, 166) = 4.17, p < .05). Moreover, the medi- Perceived Loafing and Collective Efficacy on Group Goal ated moderation model revealed a significant indirect effect through Processes and Group Performance,” Organizational Behavior perceived group entitativity (95% confidence interval: .001, .380), and Human Decision Processes, 74 (1), 62-87. which fully mediated the interactive effects of facial resemblance Platek, Steven M., Danielle M. Raines, Gordon G. Gallup Jr, and construal level on purchase likelihood. Feroze B. Mohamed, Jaime W. Thomson, Thomas E. Myers, Ivan S. Panyavin, Sarah L. Levin, Jennifer A. Davis, Ludivine Discussion C. M. Fonteyn, and Danielle R. Arigo (2004), “Reactions to This study contributes to the literature on facial resemblance Children’s Faces: Males are More Affected by Resemblance and by and large to face-based inference research. While it is well es- than Females are, and So are Their Brains,” Evolution and tablished that how individuals appear to be (e.g., facial morphology) Human Behavior, 25 (6), 394-405. can affect how they are judged by others, little research concerns per- Platek, Steven M., Samuel R. Critton, Rebecca L. Burch, David ception of groups (of individuals) purely based on facial cues. This A. Frederick, Thomas E. Myers, and Gordon G. Gallup Jr. research extends the scope of this line of research from individual- (2003), “How Much Paternal Resemblance is Enough? Sex level perception to group-level perception in the context of facial Differences in Hypothetical Investment Decisions but Not resemblance. We demonstrate in three experiments that increasing in the Detection of Resemblance,” Evolution and Human facial resemblance only enhances product purchase likelihood for Behavior, 24 (2), 81-7. consumers at high-level construals. These findings are robust across Schwarz, Norbert, and Herbert Bless (1992), “Scandals and the different operationalizations on construal level (self-measured vs. Public’s Trust in Politicians: Assimilation and Contrast message-framed vs. desirability/feasibility featured), contexts (fun- Effects,”Personality and Social Psychology Bulletin, 18 (5), draising event vs. shopping for furniture), and sample characteris- 574-79. tics (undergraduate students vs. non-student sample from mTurk). Schwarz, Norbert, and Herbert Bless (2007), “Mental Construal In practice, marketers can improve purchase behavior by morphing Processes: The Inclusion/Exclusion Model,” in Assimilation facial features of a group of salespeople and altering the advertising and Constrast in Social Psychology, ed. D. A. Stapel and J. message at an abstract high-level framing. Given the widespread us- Suls, Philadelphia, PA: Psychology Press, 119-41. age of team photos or multi-face imageries in visual media, our find- Tanner, Robin J., and Ahreum Maeng (2012), “A Tiger and a ings provide managerial implications for a wide range of audiences, President: Imperceptible Celebrity Facial Cues Influence Trust including firm advertising, employee team promotion, collaborative and Preference,” Journal of Consumer Research, 39 (4), 769- consumption, political campaign and academic collaboration. 83. Trope, Yaacov, and Nira Liberman (2003), “Temporal Construal,” REFERENCES Psychological Review, 110 (3), 403-21. Beckwith, Neil E., and Donald R. Lehmann (1975), “The Vallacher, Robin R., and Daniel M. Wegner (1989), “Levels Importance of Halo Effects in Multi-Attribute Attitude of Personal Agency: Individual Variation in Action Models,” Journal of Marketing Research, 12 (3), 265-75. Identification,”Journal of Personality and Social Psychology, DeBruine, Lisa M. (2004) “Resemblance to Self Increases the 57 (4), 660-71. Appeal of Child Faces to Both Men and Women,” Evolution Verosky, Sara C., and Alexander Todorov (2010), “Generalization and Human Behavior, 25 (3), 142-54. of Affective Learning about Faces to Perceptually Similar Förster, Jens, Nira Liberman, and Stefanie Kuschel (2008), Faces,” Psychological Science, 21 (6), 779-85. “The Effect of Global versus Local Processing Styles on Wang, Ze, Xin He, and Fan Liu (2016), “From Similitude to Assimilation versus Contrast in Social Judgment,” Journal of Success: The Effects of Facial Resemblance on Perceptions Personality and Social Psychology, 94 (4), 579-99. of Team Effectiveness,”Journal of Experimental Psychology: Hollingshead, Andrea B. (1998), “Retrieval Processes in Applied, 22 (1), 48-58. Transactive Memory Systems,” Journal of Personality and Wirtz, Jochen, and John EG Bateson (1995), “An Experimental Social Psychology, 74 (3), 659-71. Investigation of Halo Effects in Satisfaction Measures of Ip, Wai-man, Chi-yue Chiu, and Ching Wan (2006). Birds of a Service Attributes,” International Journal of Service Industry feather and birds flocking together: Physical versus behavioral Management, 6 (3), 84-102. cues may lead to trait versus goal-based group perception. Zebrowitz, Leslie A., Masako Kikuchi, and Jean-Marc Fellous Journal of Personality and Social Psychology, 90, 368-81. (2010), “Facial Resemblance to Emotions: Group Differences, Kraus, Michael W., and Serena Chen (2010), “Facial-Feature Impression Effects, and Race Stereotypes,”Journal of Resemblance Elicits the Transference Effect,”Psychological Personality and Social Psychology, 98 (2), 175-89. Science, 21 (4), 518-22. Fevered Pitch: Anxiety and Risk Avoidance in Response to Low-Pitch Ambient Sounds Michael Lowe, Georgia Institute of Technology, USA Katherine E. Loveland, Xavier University, USA Aradhna Krishna, University of Michigan, USA

EXTENDED ABSTRACT restaurant was typically well reviewed. We found a significant inter- The consumer experience is comprised of a complex and mul- action between the sound source and pitch (F(1,135)=4.87, p<.05) on tilayered soundscape that includes both scripted (e.g. spokesperson risk aversion. When the ambient sound was ascribed to strangers, par- voice, background music) and unscripted (e.g. crowd noise, the hum ticipants indicated a lower willingness to try a potentially risky food of appliances) elements. Despite the omnipresent nature of sound, if the pitch of the ambient sound was low (MLow=3.84, MModerate=4.80; relatively little research has explored how the most basic structural F(1,132)=6.72, p=.01), replicating the results of previous studies. properties of sound can impact emotional and behavioral responses However, when the ambient sound was attributed to friends our ef- among consumers. We explore how the pitch of ambient sound im- fect was attenuated (F(1,132)=.311, NS). We also found a significant pacts consumer anxiety and subsequent willingness to engage in po- interaction between the sound source conditions and pitch conditions tentially risky behaviors. Specifically, we propose that, compared to on anxiety (F(1,135)=4.29, p<.05) such that in the “strangers” con- moderate-pitched ambient sounds, low-pitched ambient sounds sig- dition participants felt less safe (more anxious) when the ambient nal a possible environmental threat resulting in an anxious response sound was low-pitched (MLow=3.59, MModerate=4.09; F(1,132)=3.56, causing heightened risk appraisal and subsequently increased risk p=.06). However, this effect was again attenuated in the friends con- avoidance. We test these relationships across six studies. dition (F(1,132)=1.11, NS). Finally, a bootstrap (N = 1000) analysis In study 1, we test whether the presence of a relatively low- shows that feelings of anxiety mediated the interaction of pitch and pitched vs. moderately-pitched ambient sound impacts the selection sound source on risk avoidance, specifically the willingness to con- of “safe” or “risky” products in a retail environment. While hidden sume a risky product (a1 x b1 = .2325, 95% C.I.=.0107 to .6292). speakers played either a 60Hz or 720Hz sine wave, shoppers at a Finally, in study 6, we test our full serial mediation model. To grocery store were presented with the option of trying a free sample this end participants listened to an ad for a risk reducing product, car of either Texas BBQ (low risk) or Bombay Curry (high risk) flavored insurance, in which we manipulated the pitch of background traffic kale chips. As expected, pitch impacted risk aversion (Chi-Square = noise (low-pitch vs. moderate-pitch vs. control). Replicating our ear- 4.05, p<.05; PROC FREQ in SAS). When the low-pitch sound was lier findings, pitch had a significant impact on risk averse behavior playing, only 21% selected the “riskier” flavor; when the higher- (F(1,168) = 5.33, p<.01). Participant were willing to pay more for pitched sound was playing, 44 % did. the advertised insurance in the low-pitch condition (M=$79.27) com- In study 2, we replicated the results from study 1 in a more con- pared to both the moderate-pitch (M=$70.10, p<.05) and the control trolled lab environment and with the addition of a control condition. conditions (M=$66.33, p<.01). Risk appraisal (a1 x b1 = -.4929; 90% Participants made a series of dichotomous choices between a certain C.I. = -1.350 to -.0597) weakly mediated the relationship between cash payoff and an unguaranteed chance at a larger payoff (Griskevi- pitch condition and risk avoidance (willingness to pay for insurance) cius et al. 2011) while speakers in the ceiling played either a relative- while anxiety alone did not (a1 x b1 = -.3544; 90% C.I. = -1.502 to ly low-pitched (80hz), a relatively higher-pitched (720hz) sine wave .0689). Anxiety did, however, mediate the relationship between pitch sound, or no sound at all (control condition). Again, pitch of the am- condition and risk appraisal (a1 x b1 = -.0429; 90% C.I. = -.1374 to bient sound had a significant impact on risk aversion (F(2,228)=4.68, -.0006) and risk appraisal mediated the relationship between anxiety and risk avoidance (a x b = -.5435; 95% C.I. = .1173to 1.458). Fi- p<.01). Participants in the low pitch condition took fewer risks (MLow- 1 1 nally, taking these steps together, there was significant serial media- Pitch=11.5) than those in both the moderate pitch (MHighPitch=14.3) and tion between pitch condition, anxiety, risk appraisal, and risk avoid- control conditions (MControl=14.5). In study 3, participants listened to 1 of 2 versions of anti- ance (a1 x d21 x b2 = -.1090; 95%C.I. = -.4609 to -.0047). smoking ad for which we varied the pitch of the background mu- Taken together, these results suggest that subtle differences in the pitch of ambient sound can have a strong influence on both emo- sic. Participants in the low pitch condition (MLowPitch=1.55) reported significantly lower intentions of smoking in the future relative to tional and behavioral responses. Specifically, compared to moderate- pitch ambient sound, low-pitch ambient sound leads to risk aversion those in the more moderately-pitched condition (MModeratePitch=2.06; F(1,128)=4.57, p<.05). Furthermore, this effect was mediated by the through anxiety and situational risk appraisal. perceived risk associated with smoking (a1 x b1 = -.2093, 95% C.I. 95% = -.5236 to -.0098). REFERENCES In study 4, we find initial evidence that our effects are driven Griskevicius, Vladas, Joshua M. Tybur, Andrew W. Delton and by anxiety. Using a word search as our D.V. we found that partici- Theresa E. Robertson (2011), “The Influence of Mortality and pants in the low-pitch condition found significantly more anxiety Socioeconomic Status on Risk and Delayed Rewards: A Life related words than participants in the moderate-pitch condition History Theory Approach,” Journal of Personality and Social Psychology, 100 (June), 1015-1026. (MLowPitch=1.26, MModeratePitch=.922; F(1,178)=4.88, p<.05) suggesting a greater sensitivity to anxiety related terms. In study 5, we provide more direct evidence that anxiety drives the relationship between the pitch of ambient sound and risk aversion and introduce a potential moderator of our effect: sound source. To this end we used a 2 (ambient sound: low pitch vs. moderate pitch) x 2 (sound source: strangers vs. friends) between subjects design. Participants imagined visiting an Asian restaurant in a neighborhood that historically had rather high crime rates, but that the food at the

Advances in Consumer Research 755 Volume 45, ©2017 Increasing Incentive Effectiveness by Linking the Incentive to a Source the Consumer Paid Into Zoe Y. Lu, University of Wisconsin - Madison, USA Robin Tanner, University of Wisconsin - Madison, USA Kurt Carlson, College of William and Mary, USA

EXTENDED ABSTRACT incentive: individual income taxes / corporate income taxes / manu- Government, employers, and marketers often use incentives to facturer concession. The dependent variable was the likelihood of motivate certain behaviors (e.g., Shapiro and Slemrod, 2003; Volpp, purchasing a new car now, elicited on a 9-point scale. Results indi- Asch, Galvin and Loewenstein, 2011). However, their effectiveness cated that participants were more likely to expedite their purchase in varies considerably. Insight as to why this might be can be found in the “individual income tax” condition (M = 6.94) than in the “corpo- multiple streams of research. Of particular relevance to our work, Ep- rate income tax” condition (M = 6.11; t(192) = 2.28, p < .05) or the ley and his collaborators (Epley and Gneezy, 2006; Epley, Mak and “car manufacturer” condition (M = 6.03; t(192) = 2.43, p < .05). A Idson, 2006) demonstrates that consumers are more likely to spend bootstrapping analysis demonstrated that that psychological owner- windfall framed as a bonus (other’s money) than as a rebate (one’s ship of the incentive fully mediated the effect of the source of incen- own money). Potentially at odds with these findings is research on tive on early purchase (β = .34, SE = .10; 95% confidence interval savings and decumulation decisions (Shu and Payne, 2015) which = [.16, .56]). finds that the more strongly people feel that Social Security benefits Study 3 examined the effectiveness of an incentive in encourag- come from money they contributed through their working life, the ing healthier diets and explored the issue of contingent vs. windfall sooner they want to claim (and presumably spend) it. incentives. The study adopted a 2 by 2 factorial design. A scenario To reconcile these findings, we note that windfalls require no described a $200 discount of a CSA membership offered by a health action to be taken in order to be received, while claiming Social insurance company. The incentive was either contingent on subscrip- Security benefits early requires a volitional decision by the retiree. tion or paid up front (in which case it could be spent elsewhere), and Building on these streams of research, we predict that consumers will was either funded by insurance premiums or donations. The depen- react to incentives differently depending on whether the incentive dent variable was the likelihood of CSA subscription, elicited on a is linked to their own money or other people’s money, and whether 5-point scale. An ANOVA revealed no main effect of the source of the incentive is conditional on their behavior, or is in the form of an the incentive (F(1, 118) = .06, p = .81), a main effect of the contin- unconditional windfall. Specifically, we hypothesize the following: gency of the incentive (F(1, 118) = 3.77, p = .05), and the predicted interaction between contingency and source of the incentive (F(1, Hypothesis 1: When receiving an incentive is contingent upon 118) = 11.67, p < .01). Specifically, when the incentive was contin- making a purchase, a consumer is more likely to gent upon CSA subscription, students were more likely to subscribe make the purchase if the incentive is framed as when the incentive was funded by insurance premiums than by dona- being funded from a source she/he has paid into. tions (Mpremiums = 3.62, Mdonations = 2.81; t(59) = 2.65, p = .01); when the incentive was not contingent upon CSA subscription, students Hypothesis 2: When receiving an incentive is not contingent were less likely to subscribe in the “premiums” condition than in the upon making a purchase, a consumer is less like- “donations” condition (Mpremiums = 2.44, Mdonations = 3.14; t(59) = 2.20, ly to make the purchase if the incentive is framed p < .05). as being funded from a source she/he has paid In summary, framing purchase-contingent incentives as being into. funded by money consumers have contributed to increases consumer’s perceived ownership over that incentive, which in turn in- In Study 1, participants imagined they were in the market for a creases their motivation to claim it. However, this effect reverses house with some flexibility on the issue of when to buy. While they for non-purchase-contingent (windfall) incentives. These findings would prefer to buy next year rather than this year, there was a fac- have potentially significant implications for a wide range of public tor that might expedite their decision, namely, a $10,000 first time policy oriented incentives that aim to motivate society-improving home buyer incentive. The experiment manipulated the source of the behaviors. For example, consider the cash-for-clunkers program, incentive: individual income taxes / corporate income taxes / seller or incentives to buy solar power systems, both would potentially be concession. The dependent variable was the likelihood of purchasing made more effective by simply framing the source of the subsidies/ a new house now, elicited on a 9-point scale. A one-way ANOVA incentives as being paid into by the target market. Similarly, in the indicated a main effect of the source of the incentive on the likeli- consumer realm, we have preliminary data that grocery store promo- hood to expedite the purchase (F(2, 201) = 4.14, p < .05). Pairwise tions are more likely to be successful if framed as being store profit comparisons showed that participants in the individual income tax funded (i.e. shoppers contributed to) than a manufacturer concession. condition were more likely to expedite the purchase (M=7.49) than participants in either the corporate tax condition (M = 6.95; t(201) = REFERENCES 2.16, p < .05) or the seller concession condition (M = 6.82; t(201) = Epley, Nicholas, Dennis Mak and Lorraine Chen Idson (2006), 2.71, p < .01). “Bonus of Rebate?: The Impact of Income Framing on Study 2 generalized the investigation to a car incentive, and Spending and Saving,” Journal of Behavioral Decision measured participants’ psychological ownership of the incentive Making, 19(3), 213-27. with three items adapted from Shu and Peck (2011). Participants —— and Ayelet Gneezy (2007), “The Framing of Financial imagined that they were in the market for a new car with some flex- Windfalls and Implications for Public Policy,” The Journal of ibility on when to buy. However, a $2,000 incentive would expire Socio-Economics, 36(1), 36-47. within a month. Again, the experiment manipulated the source of the

Shapiro, Matthew D. and Joel Slemrod (2003), “Consumer —— and Joann Peck (2011), “Psychological Ownership and Response to Tax Rebates,” The American Economic Review, Affective Reaction: Emotional Attachment Process Variables 93(1), 381-96. and the Endowment Effect,”Journal of Consumer Psychology, Shu, Suzanne and John Payne (2015), “Psychological Factors 21(4), 439-52. in Savings and Decumulation Decisions,” NA-Advances in Volpp, Kevin G., David A. Asch, Robert Galvin and George Consumer Research, 43, 136-41. Loewenstein (2011), “Redesigning Employee Health Incentives—Lessons from Behavioral Economics,” New England Journal of Medicine, 365(5), 388-90. The Risk of Autonomy: A Dual-Process Model of How Autonomy Makes the Experience Pleasurable Depending on Risk Perception Renaud Lunardo, KEDGE Business School, France Camille Saintives, INSEEC Business School, France

EXTENDED ABSTRACT and negatively related to stress. However, the relationship between Many of our daily activities are not freely chosen, and our autonomy and well-being can be complex, and the question arises freedom to choose how to behave is often constrained by rules and if the positive effects of autonomy always hold, or, said differently, norms. However, being autonomous – in other words being choiceful if being autonomous can lead to negative outcomes under specific in one’s actions (deCharms 1976; Deci and Ryan 1985) – represents circumstances. We posit that perceiving risk in the situation one has a critical condition for well-being (Ryan and Deci 2000). In this con- autonomy over can be such a circumstance. Perceived risk refers text, and unsurprisingly, research has established that consumers are to the perception of variances in the possible gains and losses that eager for autonomy (Zheng, Van Osselaer and Alba 2016), whether it could result from that particular choice (Kahneman and Tversky refers to freedom of choice (Markus and Schwartz 2010) or freedom 1979; March and Shapira 1987). Hence, people perceive a specific in behavior (Levav and Zhu 2009), and exhibit reactance where their option less risky when, controlling for expected outcomes, this op- autonomy is threatened (Brehm 1966). tion is more certain than another more probabilistic option. In this However, the assumption that being autonomous may always regard, what could be argued is that when people have autonomy in a be beneficial for consumers may be challenged. Drawing below on given situation, the perception of risk makes salient the probabilistic the literature on self-determination (Deci and Ryan 1985b) and the negative outcomes that can occur from the experience of autonomy, theory of psychological stress and coping (Lazarus and Folkman making the experience less pleasurable. Hence, we propose that risk 1984), we propose that the pleasure derived from the experience of moderates the effect of autonomy on pleasure such that when auton- autonomy varies as a function of whether consumers perceive the omy is high, a high (versus low) level of risk makes the experience situation in which they have autonomy as risky. Specifically, we pro- less (versus more) pleasurable (Hypothesis 1). pose and test a model where autonomy leads either to stress or per- According to the theory of a theory of psychological stress sonal control depending on the level of risk that they perceive in the and coping (Lazarus and Folkman, 1984), stress arises when people situation in which they exert their autonomy. evaluate a particular encounter with the environment as irrelevant to their well-being. Relying on the notion that risky situations are Theoretical Background those that are associated with uncertainty (Kahneman and Tversky 1979; March and Shapira 1987), consumers who face risk might also Autonomy encounter a situation that is less certain to make them achieve their Autonomy refers to the belief that one can freely choose how to goals as compared to a non-risky situation. In such environments that behave in a specific situation (Deci and Ryan 1985a), or the extent to are risky, autonomy can reveal immobilizing or even detrimental to which people feel free to initiate their own actions (deCharms 1976; people (Markus and Kitayama 1991) and lead to experience the situa- Deci and Ryan 1985, 2000). Autonomy is at the heart of the self-de- tion as stressful (Hypothesis 2a). On the contrary, non-risky situation termination theory (hereafter, SDT; Deci and Ryan, 1985a, b, 2000), might make the outcome more certain and lead autonomous people which posits that autonomous actions are accompanied by a sense to perceive more contingencies between their behavioral choices and of freedom and volition (Butzel and Ryan 1997). Extant research the outcomes. As such contingencies are a core component of con- consistently provided support for the beneficial effects of autonomy, trol beliefs (Seligman 1975; Skinner 1995, 1996), it may be argued mainly in the academic (Grolnick and Ryan 1987; Vallerand, For- that when risk is low, autonomy leads to a stronger perception of tier, and Guay 1997) and work domains (Kirkman and Rosen, 1999, personal control (Hypothesis 2b). Given the literature on stress and 2000). This notion of autonomy has driving positive outcomes has control seeing these affective reactions as mediators of the effects of also been observed in consumption settings. For instance, Zhang et the environment on pleasure (Folkman et al. 1986; Hui and Bateson al. (2011) found that consumers value more their goals when they 1991), both stress (Hypothesis 3a) and personal control (Hypothesis perceive that they autonomously decided to reach that goal. In the 3b) should mediate the effects of autonomy on pleasure. Further, domain of computer games, Kim, Chen and Zhang (2016) identified people who face risk might perceive the situation as threatening, and autonomy as an important condition for enjoyment. low autonomy might in this context promote defensive reactions. Be- However, this notion of autonomy as being always beneficial cause defensiveness has been shown to be less likely to occur when has recently been challenged by Chen and Sengupta (2014), who ob- a person has greater autonomy (Knee and Zuckerman 1998), people served that because ‘vice products’ are intrinsically enjoyable and who have autonomy may react more strongly to risk, leading to an therefore induce guilt, consuming a vice product leads to lower vital- increase in stress. On the contrary, when risk is low, autonomy may ity when consumers are autonomous in their decision of eating that be here beneficial to psychological well-being and leads to more per- vice product. Their results hence challenge self-determination re- sonal control. Therefore, when risk is high, autonomy may lead to search, which finds that increased decision autonomy usually height- an increase in stress and consequently on pleasure (Hypothesis 4a), ens vitality. What we propose is in line with this latter notion that while when risk is low, autonomy may increase personal control and autonomy may not always be beneficial for consumers. We develop pleasurae (Hypothesis 4b). our rationale for this notion below. Autonomy under risk Study 1 Previous research in a wide array of areas have brought evi- Procedure. A 2 × 2 between-subjects experiment manipulated dence for positive effects of autonomy. For instance, among teach- autonomy and perceived risk. Participants (N=206 students, 50% ers (Pearson and Moomaw 2005) and workers (Kalleberg, Nesheim, male) were asked to imagine themselves with a friend on vacation and Olsen 2009), autonomy is positively related to job satisfaction in Macao, where they had to rent a scooter. Pictures of the streets

Advances in Consumer Research 758 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 759 of Macao and the scooter were provided on a screen to facilitate vary across cars. Also, although people were supervised by a profes- projection in the scenario. Autonomy was manipulated following the sional instructor, some instructors were supposedly more lax than definition of autonomy as the ability to initiate behaviors (Deci and others, making autonomy subject to variation. Hence, every third Ryan 200). In the low autonomy condition, participants were told person who arrived on the parking lot after having driven a sports that they could not drive the scooter because their friend was willing car was systematically approached and asked if they would answer to drive the scooter himself. In the high autonomy condition par- some brief questions about their experience of driving the car. The ticipants were told that they could choose to drive the scooter them- same measures as in study 1 were used and adapted to the context selves. Perceived risk was then manipulated. In the high risk condi- of the study. In addition, participants rated their general risk-taking tion, participants were informed that more than 25% (versus 1% in tendency (Kim and McGill 2011). the low risk condition) of tourists who rented a scooter over the last Results. A moderated-mediation analysis was performed with 5 years had been involved in a crash. Then, participants rated how autonomy as the independent variable, stress and personal control as much pleasure they would have riding the scooter in the city, their the mediators, pleasure as the dependent variable, perceived risk as stress (Durante and Laran 2016, α=.89) and their personal control the moderator and general risk-taking tendency as a covariate. The (Cutright, Bettman, and Fitzsimons 2013; α=.83). For manipulation analysis yielded significant positive effects of both autonomy (β=.91, check purposes, participants rated their perceived autonomy (Chen p < .001) and risk (β=.73, p < .01) on pleasure, as well as their sig- and Sengupta 2014; α=.94) and how much driving the scooter in nificant interaction (β=-.15, p < .01) showing that when autonomy Macao was risky (one item). All scales were 7-point Likert scales. increases, a low level of risk leads to higher pleasure. Supporting

Results. Manipulations of autonomy (MHigh=5.13, MLow=2.44, Hypotheses 2a-b, results also revealed that neither autonomy (β=-

F(1, 204)=275.67, p < .001) and risk (MHigh=4.60,MLow=2.19 F(1, .17, p > .10) nor risk (β=.03, p > .05) had an effect on stress, but their 204)=105.31, p < .001) were successful, and the manipulation of interaction was observed (β=.10, p < .01), such that when risk is high, autonomy did not affect personal control (F(1, 204)=3.67, p > .05). being autonomous makes the experience more stressful. Personal A 2-way ANOVA revealed no effects of autonomy (p > .05) on control was not affected by risk (p > 05) but by autonomy (β=.98, p pleasure, but a negative effect of perceived risk (F (1, 204)=4.68, < .001) and their interaction (β=-.15, p < .01) such that when risk is p < .01; MLow=3.93; MHigh=3.51). Supporting Hypothesis 1, a sig- high, autonomy has no effect on personal control but has a positive nificant interaction was observed (F (1, 204)=22.01, p < .001). Pre- effect when risk is low. Hypotheses 3-4 were also supported, with cisely, in the low risk condition, autonomy had a positive effect on mediating effects of both stress (CI=-.19; -.03) and personal control pleasure (MLR=3.53; MHR=4.31, p < .01), while a negative effect was (CI=.01; .17) and similar significant indexes of moderated mediation observed in the high risk condition (MLR=3.98; MHR=3.09, p < .001). for stress (CI=-.084; -.003) and personal control (CI=-.082; -.001) Supporting Hypothesis 2a, results revealed positive effects of au- were also observed, with the same pattern of results as in study 1. For tonomy (F (1, 204)=6.89, p < .05; MLow=3.92; MHigh=4.37) and risk all analyses, general risk-taking tendency had no effects (p > .05).

(F (1, 204)=20.41, p < .001; MLow=3.78; MHigh=4.57) and, interest- Hence, study 2 replicates previous results. However, some dif- ingly, their interacting effect (F (1, 204)=3.90, p=.05), whereby in ferences pertaining to the effects of autonomy and risk on stress were the low risk condition autonomy had no effect on stress (p > .05) observed. To address these differences, we hypothesize that self- but had a positive one in the high risk condition (t=-2.97, p < .01; efficacy, or the belief that one is able to mobilize the skills that are

MLow Auton.=4.15; MHigh Auton.=4.92). Turning to Hypothesis 2b, results necessary to attain a given goal (Bandura 1982, 1997), moderates the revealed a significant positive effect of autonomy (F (1, 204)=4.71, effects of autonomy on stress, such that when autonomy increases, p < .05; MLow=3.08; MHigh=3.41) and a negative effect of risk (F (1, self-efficacy makes the experience of autonomy less stressful (Hy-

204)=38.87, p < .001; MLow=3.69; MHigh=2.78) on personal control. pothesis 5). Supporting Hypothesis 2b, a significant interaction was observed (F (1, 204)=4.25, p < .05) whereby autonomy had a positive effect on Study 3 Procedure and sample. A 2 × 2 between-subjects design was personal control when risk was low (MLow Auton.=3.35; MHigh Auton.=4.03, p < .05) but had no effect when risk was high (p > .05). conducted with autonomy (low vs. high) and risk (low vs. high) as

Then, a mediation analysis revealed significant indirect effects fixed factors. Participants (n=240, 54% female, AgeM =28.28) were of autonomy on pleasure through stress (CI=-.25; -.02) and personal asked to read a scenario where they had to imagine cooking a meal control (CI=.01; .22), supporting H3. Results from a moderated- for an evening with friends and to choose one dish to cook from a list mediation analysis where stress and personal control were included of five dishes (the dishes did not significantly differ on indulgence). as mediators and risk as a moderator revealed that the mediating ef- In the condition of high autonomy, respondents were advised that fects of stress (CI=-.35; -.01) and personal control (CI=-.37; -.01) they had the option of following a recipe or doing as they wished, were moderated by risk, but differently. Specifically and supporting while in the low autonomy condition they were specified that they Hypothesis 4, stress negatively mediated the effects of autonomy on had to follow a recipe scrupulously and that no room was left for pleasure when risk was high (CI=-.34; -.03) but not when risk was imagination. Then, in the low-risk condition, it was indicated that on low (CI=-.11; .05), while the mediating effect of personal control Internet forums consumers reported feeling that the recipe had a nice was positive in the low risk condition and (CI=.03; .37), and non- 80% chance to please. In the high risk condition, it was indicated that significant in the high risk condition (CI=-.06; .09). consumers on the forums pointed out that the recipe had only a 40% chance to please. Respondents then completed the same measures of Study 2 pleasure, stress, personal control, autonomy and risk as in previous To examine if the effects observed in study 1 replicate on the studies, as well as one of self-efficacy (Keller 2006; α=.75). field, we interviewed 122 participants (66.4% men,MAge=35.23, Results. The manipulation of autonomy and risk were success- S.D.=11.21) of car racing in the West of France during two weekends ful, and self-efficacy was not affected by the manipulation of risk in July and August 2015. We waited on the parking lot of the circuit (p > .10). As studies 1-2, the same patterns of interacting effects of for people who had just finished driving a car. Because the cars that autonomy and risk on pleasure (F (1, 236)=8.66, p < .01), stress (F people could drive vary in their power, perceived risk was likely to (1, 236)=8.01, p < .01) and personal control (F (1, 236)=7.59, p < 760 / The Risk of Autonomy: A Dual-Process Model of How Autonomy Makes the Experience Pleasurable Depending on Risk Perception

.01) were observed. Also, results yielded the same mediating effects Grolnick, Wendy S. and Richard M. Ryan (1987), “Autonomy of stress (CI=-.51; -.19) and personal control (CI=.01; .14), that were in Children’s Learning: An Experimental and Individual again moderated by risk perception (CIStress=-.80; -.16; CIControl=-.37; Difference Investigation,”Journal of Personality and Social -.02). Finally, a regression revealed that when risk is high, stress is Psychology, 52, 890–898. marginally affected by efficacy (β=-.17, p=.10) and its interaction Hui, Michael K. and John E.G. Bateson (1991), “Perceived Control with autonomy (β=-.31, p < .05), whereby for all values of self-ef- and the Effects of Crowding and Consumer Choice on the ficacy below 5.57, having autonomy results in significantly higher Service Experience,” Journal of Consumer Research, 18 (2), levels of stress than not having autonomy. Age and gender had no 174–84. effects on self-efficacy (p’s > .05). Kahneman, Daniel and Amos Tversky (1979), “Prospect Theory: An Analysis of Decision Under Risk,” Econometrica, 47 Conclusion (March), 263–291. The threes studies reported here support our hypothesis that the Kalleberg, Arne L., Torstein Nesheim, and Karen M. Olsen pleasure derived from being autonomous can vary as a function of (2009), “Is Participation Good or Bad for Workers? Effects risk perception. The extent to which a consumption experience is of Autonomy, Consultation and Teamwork on Stress among pleasurable varies according to the degrees of stress and personal Workers in Norway,” Acta Sociologica, 52 (2), 99–116. control that consumers may feel in response to their autonomy and Keller, Punam (2006), “Regulatory Focus and Efficacy of Health risk. Hence, marketers should reduce perceived risk so that autono- Messages,” Journal of Consumer Research, 33 (June), 109–14. my leads to an increase in personal control and pleasure. Kim, Sara, and Ann L. McGill (2011), “Gaming with Mr. Slot or Gaming the Slot Machine? Power, Anthropomorphism, and REFERENCES Risk Perception,” Journal of Consumer Research, 38 (June), Bandura, Albert (1982), “Self-Efficacy Mechanism in Human 94–107. Agency,” American Psychologist, 37, 122–147. Kim, Sara, Rocky Peng Chen, Ke Zhang (2016), Bandura, Albert (1997), Self-efficacy: The exercise of control, New “Anthropomorphized Helpers Undermine Autonomy and York: Freeman. Enjoyment in Computer Games,” Journal of Consumer Brehm, Jack W. (1966), A Theory of Psychological Reactance, New Research, 43 (2), 282–302. York: Academic. Kirkman, Bradley L. and Benson Rosen (1999), “Beyond Self- Butzel, Jessica S. and Richard M. Ryan (1997), “The Dynamics Management: Antecedents and Consequences of Team of Volitional Reliance: A Motivational Perspective on Empowerment,” Academy of Management Journal, 42, 58–74. Dependence, Independence, and Social Support,” In Pierce, Kirkman, Bradley L. and Benson Rosen (2000), “Powering Up Gregory R., Brian Lakey, Irwin G. Sarason, and Barbara R. Teams,” Organizational Dynamics, 28, 48–66. Sarason (Eds.), Sourcebook of Social Support and Personality, Knee, C. Raymond and Miron Zuckerman (1998), “A Nondefensive New York: Plenum Press, 49–67. Personality: Autonomy and Control as Moderators of Chen, Fangyuan and Jaideep Sengupta (2014), “Forced to Be Bad: Defensive Coping and Self-Handicapping,” Journal of The Positive Impact of Low-Autonomy Vice Consumption on Research in Personality, 32, 115-130. Consumer Vitality,” Journal of Consumer Research, 41 (4), Lazarus, Richard S. and Susan Folkman (1984), Stress, Appraisal 1089-1107 and Coping, New York: Springer. Cutright, Keisha, James R. Bettman, and Gavan J. Fitzsimons Levav, Jonathan and Rui (Juliet) Zhu (2009), “Seeking Freedom (2013), “Putting Brands in Their Place: How a Lack of Control Through Variety,” Journal of Consumer Research, 36 (4), Keeps Brands Contained,” Journal of Marketing Research, 50 600–610. (June), 365–77. March, James G. and Zur Shapira (1987), “Perspectives on Risk DeCharms, Richard (1976), Enhancing Motivation: Change in the and Risk Taking,” Management Science, 33 (11), 1404–1418. Classroom, New York: Irvington. Markus, Hazel Rose and Shinobu Kitayama (1991), “Culture Deci, Edward L. and Richard M. Ryan (1985a), Intrinsic and the Self: Implications for Cognition, Emotion, and Motivation and Self-Determination in Human Behavior, New Motivation,” Psychological review, 98 (2), 224–253. York: Plenum Press. Markus, Hazel Rose and Barry Schwartz (2010), “Does Choice Deci, Edward L. and Richard M. Ryan (1985b), “The General Mean Freedom and Well-Being?,” Journal of Consumer Causality Orientations Scale: Self-Determination in Research, 37 (2), 344–355. Personality,” Journal of Research in Personality, 19, 109–134. Pearson, Carolyn L. and William Moomaw (2005), “The Deci, Edward L. and Richard M. Ryan. (2000), “The ‘What’ Relationship between Teacher Autonomy and Stress, and ‘Why’ of Goal Pursuits: Human Needs and the Self- Work Satisfaction, Empowerment, and Professionalism,” Determination of Behavior,” Psychological Inquiry, 11, Educational Research Quarterly, 29 (1), 38–54. 227–268. Ryan, Richard M. and Edward L. Deci (2000), “SDT and the Durante, Kristina M. and Juliano Laran (2016), “The Effect of Facilitation of Intrinsic Motivation, Social Development, and Stress on Consumer Saving and Spending,” Journal of Well-Being,” American Psychologist, 55, 68–78. Marketing Research, 53 (October), 814–828. Seligman, Martin E. P. (1975). Helplessness: On Depression, Folkman, Susan, Rochard S. Lazarus, Christine Dunkel-Schetter, Development, and Death, San Francisco: Freeman. Anita DeLongis, and Rand J. Gruen (1986), “Dynamics of Skinner, Ellen A. (1995), Perceived Control, Motivation, and a Stressful Encounter: Cognitive Appraisal, Coping, and Coping, London: Sage Publication. Encounter Outcomes,” Journal of Personality and Social Skinner, Ellen A. (1996), “A Guide to Constructs of Control,” Psychology, 50 (5), 992–1003. Journal of Personality and Social Psychology, 71 (3), 549–71. Advances in Consumer Research (Volume 45) / 761

Vallerand, Robert J., Michelle S. Fortier, and Frederic Guay (1997), Zheng, Yanmei, Stijn M.J. Van Osselaer, and Joseph W. Alba “Self-Determination and Persistence in a Real-Life Setting: (2016), “Belief in Free Will: Implications for Practice and Toward a Motivational Model of High school Drop Out,” Policy,” Journal of Marketing Research, 53 (6), 1050–1064. Journal of Personality and Social Psychology, 72, 1161–1176. Zhang, Ying, Jing Xu, Zixi Jiang and Szu-chi Huang (2011) “Been There, Done That: The Impact of Effort Investment on Goal Value and Consumer Motivation,” Journal of Consumer Research, 38, 78–93. I Speak Starbucks, do You? Distinction and Inclusion on Linguistic Brand Codes Ignacio Luri, University of Arizona, USA Hope Schau, University of Arizona, USA

EXTENDED ABSTRACT baristas, and endorsed by most fans) is to discount the need for the Successful brands are active producers of cultural objects, code and the barriers to learning it without denying its complexity. rather than mere parasites that appropriate popular culture. Indeed, Among the most brand loyal customers endorsing the latter dis- brands create symbols, myths, and even prescriptive models for the course, competence in Starbucks linguistic code signals commitment way we think and the way we behave – our goals, ideals, and values. to the brand (and hence, status within the fan community). These Eavesdropping on someone ordering a drink at Starbucks or talk- fans often police the brand discourse and community in such a way ing about a sci-fi narrative brand will make obvious that brands also that lack of fluency is tolerated, if looked down upon, but opposi- change the way consumers talk. Despite the widely-accepted central- tional discourse is attacked and silenced. ity of language to culture (Crawford & Valsiner 1999), brand-specific On the other ideological end, oppositional customers, in their linguistic codes have been largely overlooked in cultural approaches own words, “make a point” of not learning or using the linguistic to consumer research. brand code while ridiculing those who do and sharing with like- Despite the lack of deliberate examination, the existence of lin- minded users their successes in defying Starbucks cultural domi- guistic codes shared and used within consumption communities and nance. Interestingly, they do so while still being Starbucks customers brand communities is widely recognized. Thomsen, Straubhaar and (many of them declared frequent customers). Bolyard (1998) use the words “jargon or group speak” to describe Amid this dialectical war, occasional and unaligned Starbucks the insider code used within cyber-communities. Schau and Muñiz customers must still manage to navigate the brand code. Lacking flu- (2002) find a “community jargon” shared in brand-related personal ency, they do so using a variety of compensatory strategies that typi- websites. Langerak et al. (2003) use the term “specialized language” cally result in negative feelings or suboptimal outcomes. for the jargon spoken within an e-brand´s virtual community. In their Fans use brand codes as a field-specific cultural capital and a ethnography of the new bikers, Schouten and McAlexander (1995) source of identity, belonging, and group status. In the case of a highly define subcultures of consumption as having unique jargons, and accessible brand like Starbucks, cultural capital is more tied to how their data illustrates several examples of “biker vernacular.” One of to consume and less to what is consumed (Holt 1998). Whereas less- the Trekker sub-communities described by Kozinets (2001) revolves than-expert customers are needed if code competence is to be a sig- around the constructed language Klingon. Muñiz and Schau (2005) nal of high status, deniers of the value of “speaking the brand” are describe a characteristic “formulaic language” within users of the a menace to status-seeking fans who have learned the code. Lack of discontinued Apple Newton. Whereas the presence of brand codes fluency is described by consumers as a source of rejection and anxi- has been repeatedly acknowledged, their relevance to consumer re- ety. Consistent with previous theory (Bourdieu and Passeron 1990, searcher theory and marketing implications have yet to be explored. Lamont and Lareau’s 1988), cultural and social exclusion prove to This study aims to fill this gap in the literature by putting the spot- be a central component of the status game. On the other hand, and light on this particular brand cultural object. deviating from previous literature on consumer resistance, we find We bring together Bourdieu’s cultural capital theories (e.g. oppositional consumers who are “anti-Starbuckese” rather than anti- Bourdieu 1986, Bourdieu & Passeron 1990), a brand culture ap- Starbucks. Consumers can be polarized not only for and against a proach (Schroeder 2009), and a consumer research reading of socio- brand (as in Thompson & Arsel 2004, Kozinets & Handelman 2004) linguistics to study the practices and discourses created by branded but for and against a brand’s cultural objects while still consuming jargon and dialects. Starbucks is our context for studying brand the brand. Etic interpretation of this form of consumer cultural resis- codes. Brand-specific linguistic codes were studied using an netno- tance is that consumers are inverting the sign of marketer-imposed graphic (Kozinets 2002) approach complemented with in-store ob- codes (Ozanne & Murray 1995) while still using them for status- servation. Brand related conversations were extracted from official seeking purposes. As Holt (2002) points out, opposition does not and fan-created websites, Twitter messages and YouTube videos. liberate from the market, but instead drives consumes to build their Findings were further informed by naturalistic observation in Star- resistance identity through the market. We extend Holt’s findings by bucks coffeehouses. adding that even opposition to a brand’s cultural objects does not Findings confirm the appropriateness of the “brand language” emancipate from the brand: knowing and denying brand cultural ob- metaphor, revealing the ubiquity of its use among consumers and jects can be two fields within the same context; two strategies in the illustrating the many ways in which linguistic brand codes mirror same status-seeking game of cultural capital. natural languages. Specifically, this study sheds light on the largely unexplored topic of how “brand literacy” (Bengtsson & Firat 2006) REFERENCES is acquired. Careful examination of Starbucks’ linguistic brand code Bengtsson, A., & Firat, F. (2006). Brand literacy: Consumers’ reveals “jargon” to be an insufficient term to refer to a code with sense-making of brand management. Advances in Consumer grammar-like properties in which issues of syntax (e.g. word order), Research Volume 33, 375-380. pragmatics (e.g. tacitness and sub-textuality), and style (e.g. redun- Bourdieu, P. (1986). The forms of capital. In Richardson, J. G. dancy) matter. Linguistic competence in “Starbuckese” is revealed as (Ed.), Handbook of theory and research for the sociology of a continuum ranging from complete lack of fluency to “native-like” education (pp. 241–258). New York: Greenwood competence. Two main opposing discourses co-exist around Star- Bourdieu, P., & Passeron, J. C. (1990). Reproduction in education, bucks’ ordering script. A wide range of oppositional and skeptical society and culture (Vol. 4). Sage. consumers see the brand code as unnecessarily cryptic, putatively Crawford, V.M., & Valsiner, J. (1999). Varieties of discursive designed to pretentiously sound “vaguely European” or simply ex- experience in psychology: Culture understood through the ert Starbucks cultural hegemony. A second discourse (as voiced by language used. Culture & Psychology, 5, 259–269.

Holt, D. B. (1998). Does cultural capital structure American Muñiz, A. M., & Schau, H. J. (2005). Religiosity in the abandoned consumption? Journal of consumer research, 25(1), 1-25. Apple Newton brand community. Journal of consumer Holt, D. B. (2002). Why do brands cause trouble? A dialectical research, 31(4), 737-747. theory of consumer culture and branding. Journal of consumer Ozanne, J. L., & Murray, J. B. (1995). Uniting critical theory research, 29(1), 70-90. and public policy to create the reflexively defiant consumer. Kozinets, R. V. (2001). Utopian enterprise: Articulating the American Behavioral Scientist, 38(4), 516-525. meanings of Star Trek’s culture of consumption. Journal of Schau, H. J., & Muñiz, A. M. (2002). Brand communities and consumer research, 28(1), 67-88. personal identities: Negotiations in cyberspace. Advances in Kozinets, R. V. (2002). The field behind the screen: Using consumer research, 29(1), 344-349. netnography for marketing research in online communities. Schouten, J. W., & McAlexander, J. H. (1995). Subcultures of Journal of marketing research, 39(1), 61-72. consumption: An ethnography of the new bikers. Journal of Kozinets, R. V., & Handelman, J. M. (2004). Adversaries of consumer research, 22(1), 43-61. consumption: Consumer movements, activism, and ideology. Schroeder, J. E. (2009). The cultural codes of branding. Marketing Journal of consumer research, 31(3), 691-704. Theory, 9(1), 123-126. Lamont, M., & Lareau, A. (1988). Cultural capital: Allusions, Thompson, C. J., & Arsel, Z. (2004). The Starbucks brandscape gaps and glissandos in recent theoretical developments. and consumers’(anticorporate) experiences of glocalization. Sociological theory, 153-168. Journal of Consumer Research, 31(3), 631-642. Langerak, Fred, Peter C. Verhoef, Peeter W. Verlegh, and Kristine Thomsen, S. R., Straubhaar, J. D., & Bolyard, D. M. (1998). de Valck (2003), “The Effect of Members’ Satisfaction with a Ethnomethodology and the study of online communities: Virtual Community on Member Participation,” ERIM Report exploring the cyber streets. Information research, 4(1), 4-1. Series Research in Management, 2003-004-MKT, Erasmus Research Institute of Management, Erasmus University Rotterdam. Increasing PSA Effectiveness: Two Routes from Self-Threat to Message Acceptance Jingjing Ma, Peking University, China Kent Grayson, Northwestern University, USA David Gal, University of Illinois at Chicago, USA

EXTENDED ABSTRACT simply thinking about a sunny day (which is not related to self-affir- The typical goal of a public service announcement (PSA) is to mation or compensation) also enhanced message acceptance. prompt viewers to behave in a more socially responsible way. Inter- In Study 3, we demonstrated conditions under which cross- estingly, many PSAs use shocking stories or visuals to illustrate the domain compensation was effective vs. ineffective in reducing self- irresponsibility of viewers’ current behaviors. However, by illustrat- discrepancy. Moreover, we showed that even when cross-domain ing how a viewer may be acting in a socially irresponsible way, such compensation was ineffective, it could still increase message ac- messages are likely to create an identity threat and thus may lead to ceptance. Specifically, participants received a threat to one of five defensive reactions rather than to more responsible behaviors. life values (family, honesty/kindness, competency, spontaneity, and To increase acceptance of threatening messages, past research physical attractiveness), then they received an essay writing task has attempted either (1) to reduce individuals’ experience of threat designed to self-affirm one of those five life values. We found that through self-affirmation (Sherman and Cohen 2006; Sherman, Nel- the cross-domain compensation effectively reduced self-discrepancy son, and Steele 2000; Steele 1988) or (2) to elevate individuals’ mood only when the threat was to the least important value and the com- (Raghunathan and Trope 2002). In this research, we highlight the re- pensation was to the most important value. However, even when the lationship between these two solutions, we identify conditions under cross-domain compensation did not effectively reduce discrepancy, it which each is more or less effective at reducing resistance to counter- still increased participants’ mood which, in turn increased their mes- attitudinal messages, and we explore the mechanism for their influ- sage acceptance. ence on message acceptance. More specifically, we propose that the Last study, we applied our findings in a filed study to show the more important a threatened identity domain is to an individual, the role of mood in improving the effectiveness of PSA. 70 male workers more seriously a person’s identity will feel threatened, and the less (generally heavy smokers) working at a construction site in China effective cross-domain self-affirmations are therefore likely to be. As were randomly assigned to one of three conditions by receiving three suggested by Tetlock et al. (2000), sacred or important values are different types of PSA messages in print: control condition (no PSA), harder to compensate or trade-off (Tetlock et al. 2000). If and when a threat condition (a threatening PSA featuring pictures of the deadly self-affirmation does not reduce an identity threat, any effect on mes- consequences of smoking), and mood elevation condition (a PSA sage acceptance might be credited to mood. featuring pictures of healthy looking families living happily). Then In study 1, participants were randomly assigned to one of three they were asked to answer five questions regarding anti-smoking conditions: threat-only, threat-compensation, and no-threat. Par- messages, gender (70 males), age (mean=38), and how many ciga- ticipants in both the threat-only and threat-compensation conditions rettes they smoke each day (mean=19). engaged in a procedure that threatened their moral identity (i.e., ad- Results showed message acceptance in the mood elevation con- mitting to online piracy behavior). Later, participants in the threat- dition was significantly higher than that in the control condition (4.08 compensation condition were asked to write an essay to self-affirm vs. 3.49; t(44) = 3.24, p < .005) and the threat condition (4.08 vs. their competence. Lastly, all participants evaluated advertisements 3.39; t(47) = 2.47, p < .05); mood in the mood elevation condition and expressed their agreement to anti-piracy messages. Results was significantly higher than that in the control condition (4.80 vs. showed that participants in the threat-only condition showed greater 3.10; t(44) = 7.83, p < .001) and the threat condition (4.80 vs. 1.83; preference for moral advertisements than those in the no-threat con- t(47) = 12.33, p < .001). A mediation analysis showed that mood dition (p < 0.05), which was consistent with past research on within- mediates the influence of different PSAs on message acceptance, the domain compensation. Contrary to what previous self-affirmation mean indirect effect from the bootstrap analysis is also significant (a research might predict, the preference for moral advertisements for × b = .18), with a 95% confidence interval excluding zero (.0384 to those in the threat-compensation condition was not significantly dif- .3765). ferent from that in the threat-only condition (p = 0.81). This result in- This research contributes to literature on self-affirmation and dicated that participants’ threatened moral identity was not effective- persuasion and have practical implications in improving the effec- ly compensated by cross-domain self-affirmation. However, relative tiveness of PSAs. to participants in the threat-only condition, participants were more willing to accept threatening messages in the no-threat condition (5.6 REFERENCE vs. 4.57; t(92) = 3.02, p < .005), and in the threat-compensation con- Sherman, David K. and Geoffrey L. Cohen (2006), “The dition (5.56 vs. 4.57; t(87) = 2.67, p < .01). This result raises a para- Psychology of Self-Defense: Self-Affirmation Theory,” dox that why individuals become more accepting of threatening mes- Advances in Experimental Social Psychology, 38, 183-242. sages when they still have discrepancy in their threatened identity. Sherman, David K., Leif D. Nelson, and Claude M. Steele (2000), Study 2 investigated the underlying mechanism for this in- “Do Messages About Health Risks Threaten the Self? creased message acceptance. In this study, we replicated our find- Increasing the Acceptance of Threatening Health Messages Via ing that cross-domain compensation increased willingness to accept Self-Affirmation,”Personality and Social Psychology Bulletin, threatening messages, but did not reduce self-discrepancy. We also 26 (9), 1046-58. showed that cross-domain compensation improved respondents’ Steele, Claude M. (1988), “The Psychology of Self-Affirmation: mood, which mediated the effect of cross-domain compensation on Sustaining the Integrity of the Self,” Advances in Experimental message acceptance. Further supporting our claim that improved Social Psychology, 21, 261-302. mood can enhance message acceptance, we showed in this study that

Raghunathan, Rajagopal and Yaacov Trope (2002), “Walking the Tetlock, Philip E., Orie V. Kristel, S. Beth Elson, Melanie C. Tightrope between Feeling Good and Being Accurate: Mood Green, and Jennifer S. Lerner (2000), “The Psychology of the as a Resource in Processing Persuasive Messages,” Journal of Unthinkable: Taboo Trade-Offs, Forbidden Base Rates, and Personality and Social Psychology, 83 (3), 510-25. Heretical Counterfactuals,” Journal of Personality and Social Psychology, 78 (5), 853-70. Towards a New Framework of Eco-Food Consumption: The Importance of Contextual Drivers Natalia Maehle, Western Norway University of Applied Sciences, Norway Cele Otnes, University of Illinois at Urbana-Champaign, USA Nina M. Iversen, BI Norwegian Business School, Norway Leif Hem, Norwegian School of Economics, Norway

EXTENDED ABSTRACT related topics discussed in the media in the US and Norway. We also Green consumer trends demonstrate that an influential group of compared the food safety regulations in both countries. aware and ethical consumers has emerged. However, what consum- We discerned the following patterns when comparing responses ers say about the importance of ethical and environmental issues of- from Norway and the US. Like previous research (e.g., Carrigan et ten differs from their actual behavior, and the share of eco-friendly al. 2004), the current study shows consumers experience an attitude- food in total consumption has remained low (Rousseau and Vranken behavior gap in both countries. Specifically, informants offer two 2013). Frequently this gap between consumers’ attitudes and their broad reasons for not purchasing eco-food, or for not buying as much actual buying behavior refers to the so-called attitude-behavior gap as they would have liked – namely, price and access. Nevertheless, (Carrigan, Szmigin, and Wright 2004). The gap can potentially be interesting differences emerge between American and Norwegian explained by the multifaceted nature of sustainable consumption. consumers when they explain the reasons for buying eco-food. While Existing studies mostly focus on individual consumer characteris- American consumers highlight the health benefits, ethical consider- tics, which means that macro and structural causes of sustainable ations and animal welfare are the most prevalent among their Nor- consumption tend to be ignored (Thørgesen 2010). For example, the wegian counterparts. Moreover, Norwegian consumers exhibit much cultural context needs to be taken into account, as it has a direct in- stronger preferences for nationally produced food, as they consider fluence on social pressure and pro-environmental value orientation Norwegian food to be similar to eco-food along the dimensions of (Kumar, Manrai, and Manrai 2017; Strizhakova and Coulter 2013). safety and ethicality. We argue these differences can be partly ex- Thørgesen (2010) argues that both macro factors (e.g., culture) and plained by contextual factors such as cultural values, differences in structural factors (e.g., political regulation, national labelling sys- media discourse, food regulations, a nation’s geographical character- tems) are more important determinants for sustainable consumption istics and provenance effects. Norwegian culture emphasizes egali- than individual-level attitudinal variables. tarianism and harmony and de-emphasizes mastery, which calls for This study looks on one particular type of sustainable consump- selfless concern for the welfare of others (Schwartz 2006). In con- tion, namely eco-food (i.e., food encompassing environmental, ethi- trast, American individualism emphasizes affective autonomy and cal and health benefits) and addresses the wider social framework mastery at the expense of harmony (Schwartz 2006). This combina- and cultural system in which this type of food is consumed. As a tion leads to justifying and encouraging egoistic self-advancement. result, we suggest a new conceptual framework that contributes a This distinction may result in a more functional, self-focused contextual dimension to our understanding of eco-food consump- approach to eco-food in the US (with a focus on health and taste tion. In this way, we extend the existing literature by suggesting that benefits), while ethical (and especially animal welfare) issues prevail consumers’ perceptions and attitudes toward eco-food are influenced in Norway. In addition, our analysis reveals that focus on issues in by five contextual drivers: (1) cultural values; (2) media discourses the media differs in the US and Norway. While Norwegian media de- (which focus on different issues in different societies); (3)- gover votes attention to animal welfare issues, there is still more emphasis nance and law (different food safety regulations, differences in trust on health issues and food safety (e.g., pesticides and hormones) in in national/stately systems); (4) geographical characteristics (e.g., the US. Our informants confirm the influence of media discourses country size, access to locally, regionally and nationally produced and refer to recent feature stories in the media while explaining their food); and (5) country-of-origin effect/provenance (preference for reasons for buying eco-food. Moreover, Norway has more restric- national vs. local food). tive food safety regulations (EcoWatch 2014), which can explain Our aim is to explore how these contextual factors explain the consumers’ trust to non-eco-food produced in Norway. Geographical differences in consumers’ perceptions and attitudes toward eco-food differences may also play a role. Norway is much smaller country in the US and Norway. Both countries are developed and industrial- than the USA and it can explain the different perceptions of prov- ized and contain a growing eco-food market. However, they differ on enance. While consumers in the USA exhibit preference for local cultural orientations and other dimensions (country size, food regula- non-eco-food, Norwegians perceive local food much broader and tions, media discourse, and so on). We can therefore expect differ- prefer all the food produced in Norway. ences in the importance of eco-food concerns. To collect the data, we This study contributes to the existing literature by bringing conducted 40 in-depth interviews in Norway and the US (twenty in attention to the contextual factors which were earlier neglected in each country). The researcher began by discussing consumers’ food the literature (Thørgesen 2010). The findings are important for both habits and their salient concerns pertaining to food purchases. The in- companies and policy makers as we highlight the importance of tak- terviewer then explained the concept of eco-food, probing informants’ ing into account contextual dimensions while promoting eco-food perceptions of the category and its salience to their purchasing and in different countries. However, this study has a number of limita- eating behavior. Moreover, the interviewer probed for contextual fac- tions and more research is necessary. We address only two countries tors influencing consumers’ perceptions of eco-food. The interview and the research can be extended to different types of cultures, e.g. procedure yielded around 520 pages of transcribed single-spaced text. more collectivistic cultures. We also do not focus on the variations The research team utilized the constant comparative method (Glaser between ethnical groups, which can be especially important in the and Strauss 1968) to identify the presence of the contextual drivers US. Moreover, there is a need to consider how a combination of discussed in the conceptual framework. In addition, the researchers product-, consumer- and context-level drivers influences eco-food conducted a media discourse analysis to identify food and eco-food consumption.

Advances in Consumer Research 766 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 767 REFERENCES Strizhakova, Yuliya and Robin A. Coulter (2013), “The ‘Green’ Carrigan, Marylyn, Isabelle Szmigin, and Joanne Wright (2004), Side of Materialism in Emerging BRIC and Developed “Shopping for a Better World? An Interpretive Study of the Markets: The Moderating Role of Global Cultural Identity,” Potential for Ethical Consumption within the Older Market,” International Journal of Research in Marketing, 30 (1), 69–82. Journal of Consumer Marketing, 21 (6), 401–417. Rousseau, Sandra and Liesbet Vranken (2013), “Green Market EcoWatch (2014), “13 Ways the EU Beats the U.S. on Food Expansion by Reducing Information Asymmetries: Evidence Safety,” http://www.ecowatch.com/13-ways-the-eu-beats-the- for Labeled Organic Food Products,” Food Policy, 40, 31–43. u-s-on-food-safety-1881850175.html. Schwartz, Shalom H. (2006), “A Theory of Cultural Value Glaser, Barney and Anselm Strauss (1968), The Discovery of Orientations: Explication and Applications,” Comparative Grounded Theory: Strategies for Qualitative Research, Sociology, 5 (2), 137–182. Chicago, IL: Aldine. Thørgesen, John (2010), “Country Differences in Sustainable Kumar, Bipul, Ajay K. Manrai, and Lalita A. Manrai (2017), Consumption: The Case of Organic Food,” Journal of “Purchasing Behaviour for Environmentally Sustainable Macromarketing, 30 (2), 171–185. Products: A Conceptual Framework and Empirical Study,” Journal of Retailing and Consumer Services, 34, 1–9. The Double-Edged Sword of Social Closeness in Gift Giving Sam J. Maglio, University of Toronto, Canada Evan Polman, University of Wisconsin-Madison, USA Alex Kaju, University of Toronto, Canada

EXTENDED ABSTRACT 294.67) and high (M = 1430.80 days, SD = 314.69) levels of quality, Consumers constantly decide what to bring into their lives, as contrasted with the expected retention from a close friend at low from people (who to date) to places (where to live) to products (what (M = 368.76 days, SD = 215.39) and high (M = 1328.60 days, SD = to buy). Beyond such what questions, any acquisition necessitates 331.19) quality levels. a subsequent, unexamined question: How long will I keep it? We Study 3 provides discriminant validity between liking of the examine this concept of retention, or the duration over which con- gift and retention. Using a high- and low-priced pen methodology sumers maintain possession of something. To probe retention, we similar to Study 1 and including a measure of liking, Study 3 found first present a series of three pilot studies that tests a straightforward that among participants asked about liking, only a significant main prediction relating retention to gift giving. Specifically, people retain effect of gift price emerged (p < .001), whereas giver distance and the gifts for a longer period of time when they come from socially closer interaction between distance and price two did not prove reliable (ps others. We then identify a condition (explicit consideration or inclu- > .4). The general conclusion appears to suggest that expensive gifts sion of gift quality information) that causes this pattern of retention are liked more than inexpensive gifts regardless of who is giving to reverse (longer retention for gifts from distant others) as a function them. Thus, it appears that consumers consider and make retention- of gift recipients’ shifting expectations for their social connection related decisions in a manner that is conceptually dissociable from with gift givers in the future. Practical implications for gift giving mere liking. as well as theoretical relevance for future research on retention are Finally, Study 4 examines the mechanism by which the addi- discussed. tion of price information seems to provide a boost in retention for In a set of three pilot studies, we document the obvious con- gifts from distant givers: An inferred expectation by recipients of inclusion that people keep gifts longer when they receive gifts from creased closeness with givers in the future. Using new gift and price socially closer (versus distant) givers. Throughout, we find longer scenarios, we asked participants not only how long they planned to retention for gifts from objectively and subjectively closer others, retain the gift but also how, if at all, they anticipated their relationship across a range of retention measures – viz. prospective retention with the gift giver would change in the future. While more increased (“How long will I keep this new item?”); retroactive retention (“How closeness was anticipated for a distant (M = 0.28, SD = 0.61) versus long did I keep a former possession?”); and a combination of the lat- close (M = -0.08, SD = 0.28) giver in the low-priced gift condition, 2 ter (“How long have I had this, and how much longer should I hold F(1, 96) = 6.15, p = .015, ηp = .06., the high-priced gift showed an onto it?”). even larger difference in anticipated increase in closeness between We then tested whether this pattern would maintain when the distant (M = 2.72, SD = 0.61) and close (M = 0.16, SD = 0.47) 2 the price of the gift is revealed. Counterintuitively, we found in a set giver, F(1, 96) = 311.09, p < .001, ηp = .76. A reliable indirect path- of four studies that such gifts are retained longer when they origi- way emerged for both the close giver (β = 26.23, SE = 15.53; 95% CI nate from distant givers. In all, when price information was present, = 6.22-73.65) and the distant giver (β = 266.64, SE = 82.90; 95% CI retention of a high-priced gift was higher when the gift was from a = 97.56-427.98). Crucially, the model estimated the index of moder- socially distant giver than from a socially close giver (Studies 1-4). ated mediation at 240.41 (SE = 75.15), and a 95% confidence inter- This result was shown to be conditionally mediated by the expecta- val for this estimate did not include zero (90.09-389.89). This sug- tion of increased future closeness with a distant (vs. a closer) gift gests that, while gift price-evoked expectations of future relationship giver (Study 4). That is, unlike gift givers (and others more gener- change predicted retention in both gift-giver conditions, the psycho- ally) that are prima facie socially close, people who stand at a greater logical process relating price to anticipated change and retention was social distance have a greater potential to move through that distance more impactful in evaluating gifts from a distant (versus close) giver. in becoming closer. In turn, these expectations for change in con- Taken together, this research emphasizes the relevance of reten- nectedness in the future informs decisions in the present. tion as a conceptual variable. It also provides an initial treatment of it Study 1 asked participants about their retention intentions for as an experimental outcome that allows for novel consumer insights gifts that were low- or high-priced ($10 vs. $75) from a socially dis- into the relationships between social closeness, price information, tant or close giver. When price information was revealed, the pat- and gift retention. tern of results was notably different than in our pilot studies: For the low-priced gift, expected retention did not differ between close (M = 105.80 days, SD = 66.26) and distant (M = 118.80 days, SD = 42.85) givers, F < 1, p > .8; but for the high-priced gift, expected retention was significantly higher when it came from a distant giver (M = 1474.60 days, SD = 414.56) compared with a close giver (M = 2 1226.40 days, SD = 457.83), F(1, 144) = 10.15, p = .002, ηp = .07. Study 2 provides evidence that price is used as a proxy for product quality in retention decisions. Study 2 tested two different prices (as in Study 1) across two different levels of gift quality (“rather low in quality – about a 2 on a 10-point scale” vs. “very high in quality – about a 9 on a 10-point scale).” When quality of the gift was revealed, expected retention was consistently higher when the gift came from a distant friend at both low (M = 504.40 days, SD =

Advances in Consumer Research 768 Volume 45, ©2017 Same Same, But Different: How Refutational Two-sided Messages Steer Ambivalent Attitudes Anuja Majmundar, University of Southern California, USA Erlinde Cornelis, San Diego State University, USA Nico Heuvinck, Lille Catholic University, France

EXTENDED ABSTRACT factorial design. Study 2 was identical to the previous one except for The question whether marketers should use one- vs. two-sided some modifications: (a) only two-sided messages were used, (b) a messages has received considerable attention in the persuasion lit- different product stimulus was used to generalize across products: a erature (e.g., Eisend, 2006), but without straight answers. On the one notebook, (c) the attributes were pretested on their importance level, hand, the inclusion of negative information in a promotional message and, finally (d) measures of negative and positive WOM were in- seems to enhance attitudes and intentions (e.g., Crowley & Hoyer, cluded. The results show significant interaction effects: in case of 1994), whereas, on the other hand, attitudes and intentions can be high ambivalence, refutational two-sided messages lead to higher derogated (Eisend, 2006). We argue that attitude ambivalence could attitudes, purchase intentions, and proportionally more (less) posi- account for these mixed results. Past research demonstrates that at- tive (negative) WOM (compared to non-refutational two-sided mes- titude ambivalence plays an important (negative) role in influencing sages). Hence, a refutation may overcome the perverse effects of attitudes and intentions (e.g., Priester & Petty, 1996). When people ambivalence in two-sided messages. are confronted with a two-sided message, this may induce feelings of Study 3 examines whether the nature of the refutation matters. evaluative conflict (i.e., attitudinal ambivalence) which may lead to Refutational statements can be based on the same or different prod- lower attitudes and purchase intentions. However, when a two-sided uct attribute levels, for example: “This laptop has a short battery message does not evoke high levels of ambivalence, we expect posi- autonomy, but on the other hand the battery charges exceptionally tive effects on message effectiveness as previously documented in fast” (same attribute) vs. “This laptop has a short battery autonomy, message sidedness literature. but on the other hand this model has many USB ports” (different The first objective of this research is to introduce attitude am- attribute). We test whether refutation on the same vs. different at- bivalence as an explanatory factor for prior inconsistent results contributes in a two-sided message amplifies attitudes, purchase inten- cerning the effectiveness of one- vs. two-sided messages (study 1). tions, and WOM. 305 participants of an online respondent panel We know that ambivalent attitudes are generally more susceptible were randomly assigned to one of six conditions in a 3 (two-sided to persuasion (van Harreveld et al., 2009). Most research focuses non-refutational vs. two-sided refutational on the same attribute vs. predominantly on the underlying reasons why (e.g., because they are two-sided refutational on a different attribute) by 2 (high vs. low am- less strong, or because they are a source of psychological discomfort bivalence) between-subjects factorial design. Study 3 was similar to which people want to resolve). Instead, our research focuses on refu- study 2 except (a) the same vs. different product attribute level refutation as a tool to get people with highly ambivalent attitudes ‘off the tational statements were manipulated by tailoring the type of product fence’ (study 2). Next, study 3 introduces refutation on the same vs. attribute(s). The results show that among high (vs. low) ambivalent different product attributes as a moderator of these effects. This work individuals, same-attribute refutations lead to significantly propor- advances the literature on ambivalence and message sidedness, while tionally more (less) positive (negative) WOM and higher purchase also providing hands-on recommendations to practitioners dealing intentions (vs. different-attribute refutations or non-refutational two- with cases of high ambivalence. sided messages). Among individuals with low (vs. high) ambiva- 137 undergraduate students were randomly assigned to one of lence, refutational two-sided messages, irrespective of whether they four conditions in a 2 (one- vs. two-sided) by 2 (high vs. low ambiva- are based on the same or different product attributes, lead to sig- lence) between-subjects factorial design. They were shown a set of nificantly lower purchase intentions and proportionally less (more) equally important attributes that described an MP3 player. Message positive (negative) WOM (vs. non-refutational two-sided messages). sidedness was manipulated by providing either only positive or both Hence, a two-sided message with a focus on the same attribute is positive and negative product attributes. Low vs. high ambivalence more persuasive while addressing highly ambivalent people. was manipulated by varying the amount of attributes, in line with This research shows that attitude ambivalence explains previ- prior research (Priester, Petty & Park, 2007). We found significant ous inconsistent results concerning two-sided messages’ effective- interaction effects for attitude towards the MP3-player and purchase ness. That is, two-sided messages only yield more positive attitudes intention. As expected, two-sided messages only yield more positive and intentions when ambivalence is low (vs. high) (study 1). How- attitudes and intentions when ambivalence is low (vs. high). Hence, ever, in study 2, we show that this perverse effect of ambivalence when designing two-sided messages, marketers should avoid creat- in two-sided messages can be resolved by including a refutation. In ing ambivalent situations, because ambivalence offsets the positive other words, a refutation can persuade ambivalent individuals. Study effects of two-sided messages on message effectiveness. 3 shows this only works if the refutation is on the same product at- Study 2 tests whether a refutational statement in a two-sided tribute-level. While it is known that ambivalent individuals evaluate message increases message effectiveness when ambivalence is high. relevant stimuli more deeply to resolve their ambivalence, further The intent of two-sided refutational messages is to first introduce research can explore whether refutation based on the same (vs. dif- an opposing view, and subsequently demonstrate why this opposing ferent) product attributes is perceived as more salient, more com- view is inferior (Allen, 1991). A non-refutational two-sided message plete, or more successful in reducing anticipated conflict, allowing merely offers both views while not refuting the opposing view in the ambivalent individuals to form stable attitudes and intentions. message. We argue that a refutation can uplift message effectiveness in ambivalent individuals. 106 MTurk participants were randomly assigned to one of four conditions in a 2 (two-sided nonrefutational vs. refutational) by 2 (high vs. low ambivalence) between-subjects

Advances in Consumer Research 769 Volume 45, ©2017 770 / Same Same, But Different: How Refutational Two-sided Messages Steer Ambivalent Attitudes REFERENCES Priester, J.R., Petty, R.E. & Park, K. (2007). Whence univalent Allen, M. (1991). Meta-analysis comparing the persuasiveness of ambivalence? From the anticipation of conflicting reactions. one-sided and two-sided messages. Western Journal of Speech Journal of Consumer Research, 34, 11-21. Communication, 55, 390-404. Van Harreveld, F., Rutjens, B.T., Rotteveel, M.R., Nordgren, Crowley, A.E. & Hoyer, W.D. (1994). An integrative framework L.F. & van der Pligt, J. (2009). Ambivalence and decisional for understanding two-sided persuasion. Journal of Consumer conflict as a cause of psychological discomfort: feeling tense Research, 20, 561-74. before jumping off the fence.Journal of Experimental Social Eisend, M. (2006). Two-sided advertising: A meta-analysis. Psychology, 45, 167-73. International Journal of Research in Marketing, 23, 187-98. Priester, J.R. & Petty, R.E. (1996). The gradual threshold model of ambivalence: Relating the positive and negative bases of attitudes to subjective ambivalence. Journal of Personality and Social Psychology, 71, 431-449. Can Packaging Color Influence Consumers’ Willingness To Pay? Evidences of the Au Naturel-Colored Packaging Effect Veronica Marozzo, University of Calabria, Italy Maria Antonietta Raimondo, University of Calabria, Italy Gaetano Miceli, University of Calabria, Italy

EXTENDED ABSTRACT porting again H1. Consistent with H3, we found a significant two- This paper introduces the concept of Au Naturel Color and in- way interaction between packaging color and product category on vestigates the relationship between packaging color and consumers’ consumers’ WTP. Planned comparisons showed that WTP for butter willingness to pay (WTP). Drawing on color theory (Giles 2000), we was not significantly different in the two color conditions, whereas define Au Naturel Color as a color that brings to mind something that WTP for rice was significantly higher forau naturel-colored packag- comes from the soil, is not dyed or not artificial, untreated or unpro- ing than for non au naturel-colored packaging. Building upon the cessed, and expresses authenticity, that is something genuine and not results of planned comparisons, we conducted the mediation analy- altered. We include shades of beige in this color domain. sis considering only data collected on rice (N = 100; 32% females; Consistent with dual process theories (e.g., Elaboration Likeli- MAge = 32.88; SDAge = 9.44) and a moderated mediation analysis hood Model - Petty and Cacioppo 1981; and Heuristic-Systematic on the whole sample (N = 198). Results further support H2. Model - Bohner et al. 1995), packaging color can be considered as In study 3 (N = 176; 52.3% females; MAge = 22.35; SDAge = a heuristic. We propose that au naturel-colored packaging (vs. non 1.88), we found a significant main effect of packaging color, sup- au naturel-colored packaging) will work as a heuristic and produce porting again H1. Consistent with H4, we found a significant two- positive associations with the product, thus leading to an increase in way interaction between packaging color and cognitive load. Planned consumers’ WTP (H1). Moreover, the use of feelings falls in the cate- comparisons showed that in the high cognitive load condition, WTP gory of heuristic processing (Chen and Chaiken 1999) and may influ- for au naturel-colored packaging was higher than WTP for non au ence consumers’ responses based on the feeling-as-information mod- naturel-colored packaging. A similar pattern was observed in the low el under peripheral route condition (Schwarz and Clore 1988). We cognitive load condition, replicating results of study 1. As predicted, maintain that the relationship between au naturel-colored packaging however, the effect of packaging color on consumers’ WTP is sig- and consumers’ WTP is explained by an emotion-based mechanism. nificantly larger in the high cognitive load condition than in the low We expect that au naturel-colored packaging (vs. non au naturel- cognitive load condition. colored packaging) will evoke low arousal emotions, mainly con- In study 4 (N = 212; 50% females; MAge = 35.86; SDAge = sisting in sensations of tranquillity, serenity and relief, that we call 11.56), we expect to find the packaging color effect on WTP irre- safety emotions. In turn, such kind of sensations will produce higher spective of the fit between product color and packaging color. Re- consumers’ WTP (H2). We also propose that perceived dangerous- sults showed a significant main effect of packaging color, support- ness for consumers’ health of product category moderates the effect ing again H1. The interaction effect of packaging color and product of au naturel-colored packaging on consumers’ WTP. When facing a color was non-significant. Planned comparisons showed that WTP product category featuring low dangerousness consumers keep using for rice was significantly higher in the au naturel-colored packag- color packaging as a heuristic. Under such conditions, the positive ing condition than in the orange-colored packaging condition. The effect of au naturel color on consumers’ WTP will hold. Instead, the same pattern holds for carrots. Thus, our results can be explained by effect disappears for product categories featuring high dangerous- the characteristics of au naturel-colored packaging, but not by the fit ness for which heuristics are no more relevant (H3). Additionally, between product color and packaging color. the effect of au naturel-colored packaging on consumers’ WTP may Our research provides deeper insights on how consumers per- depend on cognitive load. Indeed, under conditions of high (vs. low) ceive packaging featuring products featuring au naturel-colors, cognitive load such effect may be stronger because consumers have showing that they associate to these hues of colors specific features fewer cognitive resources available which limit their ability to pro- that positively influence their WTP. Based on our results, public cess information and prompt them to use heuristics. (H4). policy makers may favour an “au naturel color” (vs. non au naturel We first validated the proposed definition of Au Naturel Color. color) system only for packaging of product categories that really Results of two preliminary studies (A and B) suggested that seven reflect the consumers’ associations in terms of genuineness and non- items aimed at measuring the characteristics of au naturel color con- artificial features. If the au naturel color system becomes a common verged onto a common latent factor, allowing us to obtain an overall practice, adopting an ethical perspective and a social fair conduct, score of Au Naturel Color Scale (ANCS). Additionally, findings al- food manufacturers may use the au naturel color system (being a lowed us to conclude that shades of beige reflect the characteristics heuristic) to communicate the value of their offerings and distinctive- of the au naturel color better than other colors in a large majority of ness from the competitors. product categories. Study B also allowed us to select stimuli for the subsequent experimental studies. Then we tested our hypotheses in REFERENCES four experimental studies. Bohner, Gerd, Gordon B. Moskowitz, and Shelly Chaiken (1995). In study 1 (N = 97; 55.1% females; ModeAge = 36-55), we “The Interplay of Heuristic and Systematic Processing of found that respondents are more willing to pay for the au naturel- Social Information,” European Review of Social Psychology, colored packaging than for the non au naturel-colored packaging, 6, 33-68. providing support to H1. A mediation analysis (Preacher and Hayes, Chen, Serena and Shelly Chaiken (1999). “Motivated Heuristic and 2004) allowed us to support the idea that safety emotions mediate the Systematic Processing,” Psychological Inquiry, 10, 44-49. effect of packaging color on willingness to pay (H2). Giles, Geoff A. (2000). Design and technology of packaging In study 2 (N = 198; 38.9% females; Mage = 34.78; SDAge = decoration for the consumer market (Vol. 1). Boca Raton, FL: 10.67), we found a significant main effect of packaging color, sup- CRC Press.

Petty, Richard E. and John T. Cacioppo (1981). Attitudes and Schwarz, Norbert and Gerald L. Clore (1988). “How do I feel Persuasion: Classic and Contemporary Approaches. Dubuque, about it? The informative function of affective states,”Affect, IA: Wm. C. Brown. cognition, and Social Behavior, 44-62. When High Failure, Risky Technology Leads to Market Expansion: The Case of the Fertility Services Market Laetitia Mimoun, HEC Paris, France Lez Trujillo Torres, University of Illinois at Chicago, USA Francesca Sobande, University of Dundee, UK

EXTENDED ABSTRACT Taking a meso-level approach to market expansion driven by This paper explores how risky innovation in the fertility servic- high-risk technological innovations, we seek to address the outlined es market impacts legitimation dynamics and marketing expansion. research gaps, by analyzing the marketing of fertility services in the Taking institutional theory as an enabling theory, we introduce scien- United Kingdom (UK) and the United States (US). Innovations in tific legitimacy as key pillar in the legitimation of risky, biomedical Assisted Reproductive Technologies (ART) have revolutionized fer- technologies. Prior research identifies several actors who play key tility services and how reproductive problems are solved, and have roles in the legitimation process of consumption practices. For ex- facilitated the quest for parenthood for various consumer demograph- ample, Humphreys (2010) highlights how casino managers achieve ics. We address two research questions: How and why does techno- a sense of legitimacy by successfully navigating regulatory, sociological innovation impact the legitimation of consumption? Under cognitive, and normative conventions that underpin the industry. Ad- what circumstances does this change in legitimacy dynamics result ditionally, Giesler’s (2012) analysis of Botox marketplace activity in market expansion? Qualitative methods and automated con- illuminates how brands strategically engage with other stakeholders tent analysis were used to analyze a set of visual and textual data, to facilitate the legitimation of new markets. Three main legitima- which includes articles from two newspapers (The Guardian and The cy-driving actors emerge in the literature: decentralized groups of New York Times) and three magazines (Elle, Cosmo, and New Scien- consumers (e.g. Martin and Schouten 2014; Scaraboto and Fischer tist), as well as marketing material from US and UK fertility clinics 2013), centralized marketplace actors such as a brand (e.g. Giesler and related health organizations. 2008; Humphreys 2010a), and news media (e.g. Humphreys and We find that decentralized technology innovation which re- Thompson 2014). lies on scientific legitimacy impacts normative legitimacy (parent- Whilst the aforementioned research has significantly advanced hood norm) and socio-affective legitimacy (in contrast to socio- our understanding of market evolution and legitimating processes, cognitive), while regulatory legitimacy plays a secondary role. The several questions remain. First, Martin and Schouten (2014) empha- socio-affective mechanism is fueled by how media and- market size that legitimation efforts include “a process of multiple transla- ers frame mythologized parenthood and time scarcity in order to tions wherein consumers mobilize human and nonhuman actors to minimize implications of risk failure. Last, despite high probability co-constitute products, practices, and infrastructure” (p.855). There- of technology failure, this market has expanded rather than con- fore, our research considers how other types of nonhuman actors tracted over time. From our frame analysis, we see that, while ART can be central in legitimation process, especially ones that originate were heavily adopted by infertile couples and single women when outside the control of marketplace players (e.g. firms or managers). it entered the mainstream, in recent years the use of ART has been Studies on role of technological innovation are limited. For instance, expanded to individuals seeking fertility preservation from cancer Giesler (2012) focuses on a technology-centered brand and empha- treatment, gay couples, couples in need of gestational surrogacy, and sizes that market creation and development evolves with consum- single, young women seeking egg preservation to name a few. ers’ perceptions of technology. We argue that technological innova- Our research contributes to literature on institutionalization, tion itself can substantially affect legitimacy dynamics concerning by showing that technology innovation is a distinct determinant of consumption practices. consumption-related legitimacy, whilst differentiating between de- Furthermore, previous related studies often focus on empirical centralized and centralized technological innovation. Finally, we link contexts in which producers and consumers are relatively clearly de- scientific legitimacy to the emotional facet of social legitimacy, and fined and the risk of consumption is minimal or predictable. Thus, its influence on consumption in a market characterized by moderate there is an impetus to examine issues pertaining to the legitimation success in helping consumers reach their goal of parenthood. Lastly, of consumption processes, in relation to contexts where the subject we hope that this work may have implications for providers looking positions of consumer and producer may shift, and where consump- to promote services in contexts where high hopes and high risk sur- tion risks are uncertain or the risk of failure exceed those of success round complex technologies, which may include a variety of medical (Fischer, Otnes, and Tuncay 2007). services. Existing work which explores the part played by communication technologies as intermediaries of legitimation processes, has fo- REFERENCES cused on the results of centralized industry efforts, such as within one Fischer, Eileen, Cele C. Otnes, and Linda Tuncay (2007), “Pursuing industry, brand, or a set of managerial or expert systems (Humphreys parenthood: Integrating cultural and cognitive perspectives and Thompson 2014). We advance that research is needed to analyze on persistent goal striving,” Journal of Consumer Research, the role of communication technologies, including mass-media, in 34(4), 425-440. selecting and disseminating information within decentralized mar- Giesler, Markus (2008), “Conflict and compromise: drama in kets to a variety of consumer segments. This is important because marketplace evolution,” Journal of Consumer Research, 34(6), in decentralized markets, news media coverage can widely advance 739-753. technological progress and reflect the type of consumers adopting Giesler, Markus (2012), “How doppelgänger brand images these innovations. This can help us to distinguish legitimation efforts influence the market creation process: Longitudinal insights from centralized industries, versus those stemming from decentral- from the rise of Botox cosmetic,” Journal of Marketing, 76(6), ized industries. 55-68.

Humphreys, Ashlee (2010a), “Megamarketing: The creation of Martin, Diane M., and John W. Schouten (2014), “Consumption- markets as a social process,” Journal of Marketing, 74(2), driven market emergence,” Journal of Consumer Research, 1-19. 40(5), 855-870. Humphreys, Ashlee, and Craig J. Thompson (2014), “Branding Scaraboto, Daiane, and Eileen Fischer (2013), “Frustrated disaster: Reestablishing trust through the ideological fatshionistas: An institutional theory perspective on consumer containment of systemic risk anxieties,” Journal of Consumer quests for greater choice in mainstream markets,” Journal of Research, 41(4), 877-910. Consumer Research, 39(6), 1234-1257. Consuming the Inauthentic as a Self-Authenticating Act in Insa-dong, Seoul Hyun Jeong Min, Black Hills State University, USA

EXTENDED ABSTRACT Korean national identity regardless of whether they enjoy or criticize As authenticating acts and authoritative performances (Arnould the offered commodities and experiences. Rather than mindless ac- and Price 2000), tourism has provided researchers with fertile ground ceptance and enjoyment, criticism on the authenticity of the place and to examine the pursuit of authenticity through consumption. Mac- the offered commodities may engage these consumers more strongly Cannell (1973) indicates that modern tourists, who are concerned in the matter of whether the commodity/service is truly Korean. In about the shallowness and inauthenticity of their lives, are motivated this process, Korean identity may be reproduced even though they by the hope of discovering and experiencing authentic cultures. Al- are not directly involved in actual consumption of such inauthentic though—as Cohen (1988) suggests—authenticity is a socially nego- commodities. Therefore, while criticizing or talking about national tiable concept perceived by individuals based on their own personal tradition and its authenticity, nations can be discursively constituted experiences and expectations (Leigh, Peters, and Shelton 2006), (Fox and Miller-Idriss 2008). prior research has established that objects and experiences that are The primary data for this research project were collected in 2013 perceived inauthentic may not have the ability to authenticate one’s through qualitative interviews with 24 native Korean consumers who individual and communal identity. However, the present study at- had prior experiences in visiting Insa-dong. Materials and photos ob- tempts to put on the weight to the other side by showing that through tained since 2004 through multiple observations in the location were cynicism, perceived inauthenticity could authenticate consumers’ used as supplement data for our project. Other supplementary data individual and communal identities, based on an examination of con- include websites of Insa-dong-related entities, government promo- sumption in the traditional cultural district of Insa-dong in Seoul, tional materials, newspaper articles, documentaries about Insa-dong, South Korea. Insa-dong is an ideal empirical setting because about movies about Korean tradition, among others. Data analysis entailed 100,000 domestic and foreign tourists visit the district every day an iterative process of coding and categorizing. Provisional under- (Insa Traditional Culture Preservation Association 2012), but many standings of consumers’ experiences were formed, challenged, and Koreans often criticize it as inauthentic and overly commercialized. revised in multiple readings of the individual transcripts and the data While early studies on cynicism tend to view it as a psycho- set as a whole in conjunction with emerging themes, including cyni- logical defect that needs to be corrected (Fleming and Spicer 2003), cism on Insa-dong and the relation between the consumption of Insa- more recent research shows that cynicism is not necessarily unethical dong and the consumer self. and does not lead to negative outcomes for organizations, employ- In contrast to a typical description of the place as “a museum ees, and consumers, emphasizing its role in social critique (Karfakis alive with the lives, histories, and cultures of Korean people” (Insa and Kokkinidis 2011) and consumer resistance to the consumer- Traditional Culture Preservation Association 2012), the research par- ist marketplace (Mikkonen, Moisander, and Firat 2011). However, ticipants perceived Insa-dong as an inauthentic representation of Ko- both negative and positive cynicism often have an effect on indi- rean traditional culture despite its official christening asa traditional viduals’ identity construction as protection against aggressive mar- cultural district. Data analysis demonstrates that despite its contested keting practices (Helm 2004; Odou and de Pechpeyrou 2010) and authenticity, however, consuming the place brand of Insa-dong au- overwhelming managerial controls (Casey 1995; Kunda 1992). As thenticates the individual consumer self and reproduces a national seen in Mikkonen et al. (2011), cynicism is also part of the consumer identity as Korean through engaged cynicism on commercialization, identity project by allowing people to take the moral high ground governmental control, and the static view of tradition that the gov- to disillusioned members of society and organizations. However, it ernment and businesses in the district take in constructing the place. does not necessarily lead to reforming actions that help change soci- Through our analysis, Korean nationalism, which incorporates ety or organizations (Bloom 2008) because while cynical individu- both banal and hot nationalism based on the historical trajectory of als may dis-identify with cultural prescriptions of a given society or Koreans who experienced both brutal colonialism and rapid econom- organization they still perform them (Fleming and Spicer 2003), and, ic development after the Korean War, has emerged as a condition that arguably, one such case in modern society is nationalism. makes consumers engage in the consumption of the place despite While nation and nationalism have long been discussed as a their perceived inauthenticity of the themed place. This reinforces contingent outcome of modernity (Gellner 1983) based on artificial the association of Insa-dong and Korean tradition and ultimately the invention of tradition (Hobsbawm 1983), recent studies on nation- reproduction of Korean national identity for the consumers of the alism examine everyday reproduction of nationhood. They high- place. light that the seemingly declining importance of national identity, By examining the consumption practices in a traditional cultural as commonly claimed by postmodernists, may be based on the fact district of Insa-dong, this study has illustrated the role of cynicism that people take national identity for granted in their everyday life. as an important mechanism to authenticate the self for modern con- Billig (1995) calls this taken for granted national identity banal na- sumers who often face inauthentic market offerings. This study also tionalism. In their everyday life, the citizens of a nation are reminded shows that socio-cultural ideology such as nationalism is an impor- of their national place in a world of nations, but this reminding is tant condition for consumers’ ongoing engagement in inauthenticity. so familiar that it is not consciously registered as a reminding, thus reproducing embodied national identity (Billig 1995). For instance, REFERENCES nations are reproduced through ordinary people’s mundane practic- Arnould, Eric J and Linda L Price (2000), “Authenticating Acts es (Fox and Miller-Idriss 2008) such as consumption of consumer and Authoritative Performances: Questing for Self and goods (Belk and Paun 1995; Douglas and Isherwood 1979). In this Community,” in The Whys of Consumption: Contemporary vein, consumption practices in Insa-dong may play an important role Perspectives on Consumer Motives, Goals, and Desires, ed. S. in reproducing Korean identity. Korean tradition-themed commodi- Ratneshwar, David Glen Mick and Cynthia Huffman, London: ties and experiences in the district may flag or remind consumers of Routledge, 140-63.

Belk, Russell W. and Janeen Arnold Costa (1998), “The Mountain Hobsbawm, Eric J. (1983), “Introduction: Inventing Traditions,” in Man Myth: A Contemporary Consuming Fantasy,” Journal of The Invention of Tradition, ed. Eric J. Hobsbawm and Terence Consumer Research, 25 (3), 218-40. Ranger, Cambridge: Cambridge University Press, 1-14. Belk, Russell W. and Magda Paun (1995), “Ethnicity and Insa Traditional Culture Preservation Association (2012), Consumption in Romania,” in Marketing in a Multicultural “Welcome Message,” http://www.hiinsa.com/eng/insadong/ World, ed. Janeen Arnold Costa and Gary J. Bamossy, insadong1.asp. Thousand oaks, CA: Sage, 180-208. Karfakis, Nikos and George Kokkinidis (2011), “Rethinking Billig, Michael (1995), Banal Nationalism, London: Sage. Cynicism: Parrhesiastic Practices in Contempory Workplaces,” Bloom, Peter (2008), “Capitalisms Cynical Leviathan: Culture and Organization, 17 (4), 329-45. Cynicism,Totalitarianism, and Hobbes in Modern Capitalist Kunda, Gideon (1992), Engineering Culture: Control and Regulation,” International Journal of Zizek Studies, 2 (1), Commitment in a High-Tech Corporation, Philadelphia, PA: 1-30. Temple University Press. Casey, Catherine (1995), Work, Self and Society: After Leigh, Thomas W., Cara Peters, and Jeremy Shelton (2006), Industrialism, London: Sage. “The Consumer Quest for Authenticity: The Multiplicity of Cohen, Erik (1988), “Authenticity and Commoditization in Meanings within the Mg Subculture of Consumption,” Journal Tourism,” Vol. 15, 371. of the Academy of Marketing Science, 34 (4), 481-93. Douglas, Mary and Baron C. Isherwood (1979), The World of MacCannell, Dean (1973), “Staged Authenticity: Arrangements of Goods, New York: Basic Books. Social Space in Tourist Settings,” The American Journal of Fleming, Peter and André Spicer (2003), “Working at a Cynical Sociology, 79 (3), 589-603. Distance: Implications for Power, Subjectivity and Mikkonen, Ilona, Johanna Moisander, and A. Fuat Firat (2011), Resistance,” Organization, 10 (1), 157-79. “Cynical Identity Project as Consumer Resistance-the Scrooge Fox, Jon E. and Cynthia Miller-Idriss (2008), “Everyday as a Social Critic?,” Consumption, Markets & Culture, 14 (1), Nationhood,” Ethnicities, 8 (4), 536-63. 99-116. Gellner, Ernest (1983), Nations and Nationalism, Ithaca, NY: Odou, Philippe and Pauline de Pechpeyrou (2010), “Consumer Cornell University Press. Cynicism: From Resistance to Anti-Consumption in a Helm, Amanda (2004), “Cynics and Skeptics: Consumer Disenchanted World?,” European Journal of Marketing, 45 Dispositional Trust,” in Advances in Consumer Research (11/12), 1799-808. Vol. 31, ed. Barbara E. Kahn Kahn and Mary Frances Luce, Valdosta, GA: Association for Consumer Research, 345-51. Form, Function, and Fluency: Surprising Effects of Product Presentation (Dis)fluency on Attribute Evaluations James A. Mourey, DePaul University, USA Aparna Labroo, Northwestern University, USA

EXTENDED ABSTRACT presentation of two very different products should highlight their dif- Relative comparison suggests that presenting very different (v. ferences and result in more extreme evaluations, particularly on the very similar) items sequentially should accentuate the stark differ- dimension on which the stimuli differ, the metacognitive experiences ences among the items such that a hedonic, pleasurable item should of fluency and disfluency that arise from homogeneous and hetero- seem much more pleasurable following the presentation of a func- geneous product presentation yield more extreme and more conser- tional, utilitarian item as opposed to following the presentation of a vative evaluations, respectively. Thus, prior rational theorizing that hedonic item. suggests heterogeneous presentation should lead to greater contrast Indeed, previous research, such as the two-factor model on as- in evaluations does not explain the muted findings found within the similation and contrast (Meyers-Levy and Sternthal 1993) would heterogeneous presentation evaluations herein. More work can be suggest that stimuli that have little overlap should produce contrast done to explore the relationship between the prior research and the effects granted plenty of cognitive resources are available. Other- current findings, specifically with respect to when consumers are wise, assimilation is predicted. Other research exploring the juxta- more likely to rely on rationality versus their metacognitive experi- position of two or more products in a market share similar effects of ence. dominance (Huber, Payne, and Puto 1982), contrasted preferences, With respect to practice, the presentation of product options and and even surprising consequence on choice (Wilcox et al. 2009). varieties has been an important cornerstone of marketing communi- However, the prior research focused primarily on the rational, cations for decades but plays an even more important role in online logical considerations of product comparisons and less, if at all, on markets where customers can scroll through hundreds to thousands the metacognitive experience of fluency that can arise from observ- of products on a website. Whether it makes sense to feature similar ing products in choice contexts. In the current research, we pro- products or different products in sequences may depend on the mar- pose that the presentation of heterogeneous or homogeneous sets keting objective of a company, as there may be instances in which of products can affect an individual’s metacognitive experience of differences should be highlighted (e.g., comparing with a competi- fluency and, in turn, affect the individual’s evaluation of a product. tor’s products) or minimized (e.g., offering alternatives for an out- More specifically, we propose that the presentation of homogeneous of-stock option). Equally important, understanding the mindset of products in a given set will be experienced as fluent and, as such, consumers matters, as customers engaging in more rational, delib- result in more heuristic, extreme evaluations. The presentation of erate decision making may exhibit contrast and assimilation effects heterogeneous products (i.e., varying on even just one dimension) consistent with the prior literature whereas other customers engaging in a given set will be experienced as disfluent and, as such, result in in less rational processing may be more likely to rely on the metacog- more deliberative, conservative evaluations, thereby attenuating the nitive experiences shown here. evaluations of the products. To test our theory, we conducted three studies using products REFERENCES selected from the famous SkyMall magazine each pretested to be Dechêne, Alice, Stahl, Christoph, Hansen, Jochim, and Michaela either utilitarian or hedonic. In study 1 (n = 122), utilitarian products Wänke (2010), “The Truth About the Truth: A Meta- presented with other utilitarian products (i.e., fluent) were rated as Anlaytic Review of the Truth Effect,”Personality and Social significantly more useful/functional than when they were presented Psychology Review, 14 (2), 238-257. with hedonic products (i.e., disfluent). In study 2 (n = 125), hedonic Huber, Joel, Payne, John. W., and Puto, Christopher (1982), products presented with other hedonic products (i.e., fluent) were “Adding Asymmetrically Dominated Alternatives: Violations rated as significantly more fun/funny than when they were presented of Regularity and the Similarity Hypothesis,” Journal of with utilitarian products (i.e., disfluent). These findings are con- Consumer Research, 9 (1), 90-98. trary to what the two-factor theory would predict regarding stimuli, Lynch, John G., Chakravarti, Dipankar, and Anusree Mitra (1991), overlap, and the likelihood for contrast or assimilation. In study 3 “Contrast Effects in Consumer Judgments: Changes in Mental (n = 118), we use products within the same product category (i.e., Representations or in the Anchoring of Rating Scales,” Journal shirts) to rule out the possibility that effects in the first two studies of Consumer Behavior, 18, 284-297. were due to product variation as opposed to the varying utilitarian/ Meyers-Levy, Joan, and Sternthal, Brian (1993), “A Two-Factor hedonic attributes of the products. Study 3 replicated the findings of Explanation of Assimilation and Contrast Effects,”Journal of the prior studies by showing that utilitarian shirts (i.e., dress shirts) Marketing Research, 30 (3), 359-368. are perceived to be more useful/functional when presented in a series Tversky, Amos (1972), “Elimination by Aspects: A Theory of with other utilitarian shirts, hedonic shirts (e.g., a Hawaiian shirt) are Choice,” Psychological Review, 79 (July), 291-299. perceived to be more fun/funny when presented in a series with other Whittlesea, B.W.A., and J.P. Leboe (2003), “Two Fluency hedonic shirts, and the heterogeneous presentation of shirts mitigates Heuristics (and How to Tell Them Apart),” Journal of Memory their usefulness/functionality (for utilitarian shirts) and their fun/fun- and Language, 49, 62-79. niness (for hedonic shirts). Wilcox, Keith, Beth Vallen, Lauren Block, and Gavan J. Fitzsimons With respect to theory, the studies contribute to the prior re- (2009), “Vicarious Goal Fulfillment: When the Mere search on contrast effects, the two-factor model on assimilation and Presence of a Healthy Option Leads to an Ironically Indulgent contrast, and other, more rational approaches to understanding com- Decision,” Journal of Consumer Research, 36 (3), 380-93. parisons between stimuli by adding a metacognitive, “feelings as information” perspective to the literature. Although rationally the

Advances in Consumer Research 777 Volume 45, ©2017 When Do Limited Time Offers Work? A Reactance Perspective Ashesh Mukherjee, McGill University, USA Seung Yun Lee, Konkuk University, South Korea Andrew D. Gershoff, University of Texas at Austin, USA

EXTENDED ABSTRACT more, reduction in purchase intent due to the reactance manipulation Marketers often use limited-time offers to attract customers. For was found to be mediated by the manipulation check for reactance. example, retailers make offers that are for “one week only”; flash Study 3 tested the hypothesis that decreasing the time to ex- sales websites make offers stating “today’s deal” or “deal expires piry will increase reactance to limited time offers, and thus reduce in one hour”; and online shopping portals make limited-time offers the positive effect of limited time offers on product evaluation. One with a countdown clock ticking down the remaining time. Given the hundred and thirteen student participants evaluated an advertisement popularity of limited time offers, it is important to understand when for a spa package with or without a limited time offer, featuring short these offers do or do not work in terms of improving product evalu- or long expiry times. Consistent with a reactance mechanism, the ation. Past research has found that putting any kind of restriction limited time offer increased purchase intent for longer expiry times on offers – such as time restriction or quantity restriction – tends to such as one week and one day, but not for shorter expiry times such increase product evaluation because restrictions prompt consumers as one hour and one minute. Finally, study 4 investigated whether to make inferences of product value (Inman et al. 1997; Parker and negative reactance responses to limited time offers can be overcome Lehman 2011). The present research extends this literature by apply- by reminding consumers of future regret if they miss the offer. Two ing the theoretical lens of reactance to limited time offers, and show- hundred and fifty nine student participants evaluated an online adver- ing that the positive effect of limited time offers on product evalua- tisement for a backpack with a short-horizon limited time offer. The tion is moderated by factors related to reactance. advertisement manipulated the presence versus absence of a limited Reactance has been defined as the motivational state aroused time offer, as well as the presence versus absence of an anticipated when individuals perceive threats to their freedom of action (Brehm regret reminder. Results showed that the presence of an anticipated 1966; Levav and Zhu 2009). We propose that limited time offers can regret reminder was sufficient to overcome reactance responses and prompt reactance when consumers perceive that their freedom to buy restore a positive effect of the limited time offer on purchase intent. the product whenever they want is being overly restricted. For ex- In summary, our findings make three theoretical contributions ample, compared to the offer “20% off regular price” which has no to the literature on limited time offers. First, we identify a new reac- time restriction, a limited-time offer stating “20% off regular price, tance mechanism underlying the effect of limited time offers on prod- deal expires in one hour” gives consumers lesser freedom regarding uct evaluation. Second, we identify new reactance-based moderators when the offer can be accepted. We suggest that such restrictions of the effect of limited time offers on product evaluation, i.e., trait on one’s freedom of choice timing can prompt reactance, which in reactance, an experimental manipulation of reactance, and time to turn can reduce the positive effect of limited time offers on product expiry of the offer. Third, we show that reminders of future regret can evaluation. We verify this proposed reactance mechanism in series overcome reactance and restore the positive effect of limited-time of- of four studies by showing that a set of factors related to reactance – fers on product evaluation. Methodologically, we present convergent trait reactance, an experimental manipulation of reactance, and time evidence across four studies using different product categories and to expiry of the offer – moderate the positive effect of limited time measures of product evaluation. Managerially, our results give new offers on product evaluation. We also show that negative reactance insight into when the widely used promotional technique of limited responses to limited time offers can be overcome by reminding con- time offer is likely to be effective for increasing product evaluation. sumers of future regret if they miss the offer. Study 1 tested the hypothesis that individuals high in trait re- REFERENCES actance are likely to experience greater reactance to limited time of- Aggarwal, Praveen, Sung Youl Jun, and Jong Ho Huh (2011), fers than those low in trait reactance, which in turn should reduce “Scarcity Messages,” Journal of Advertising, 40 (3), 19-30. the positive effect of limited time offers. One hundred and eighty Aguirre-Rodriguez, Alexandra (2013), “The Effect of Consumer student participants read an advertisement for a new brand of digital Persuasion Knowledge on Scarcity Appeal Persuasiveness,” camera which manipulated the presence versus absence of a limited Journal of Advertising, 42 (4), 371-79. time offer by including or omitting the phrase “one day only, buy Aiken, Leona S. and Stephen G. West (1991), Multiple Regression: now!” Participants then indicated their intent to purchase the adver- Testing and Interpreting Interactions, Thousand Oaks, CA: tised camera, and responded to a multi-item scale for trait reactance Sage. (Hong and Faedda 1996). Consistent with a reactance mechanism, Amaldoss, Wilfred and Sanjay Jain (2005), “Pricing of Conspicuous spotlight analysis showed that the presence (versus absence) of the Goods: A Competitive Analysis of Social Effects,”Journal of limited time offer increased purchase intent for low trait reactance re- Marketing Research, 42 (1), 30-42. spondents, but not for high trait reactance respondents. Study 2 tested Brehm, Jack Williams (1966), A Theory of Psychological the hypothesis that experimentally increasing reactance will reduce Reactance, Cambridge, MA: Academic. the positive effect of limited time offers on product evaluation. Two Castro, Iana A., Andrea C. Morales, and Stephen M. Nowlis hundred and forty seven participants from Amazon Mechanical Turk (2013), “The Influence of Disorganized Shelf Displays and evaluated an advertisement for a pair of headphones which manip- Limited Product Quantity on Consumer Purchase,” Journal of ulated the presence versus absence of a limited time offer, as well Marketing, 77 (4), 118-33. as reactance to the limited time offer. Consistent with a reactance Dubois, David, Derek D. Rucker, and Adam D. Galinsky (2016), mechanism, the limited time offer increased purchase intent in the “Dynamics of Communicator and Audience Power: The control condition but not in the high reactance condition. Further- Persuasiveness of Competence versus Warmth,” Journal of Consumer Research, 43 (1), 68-85.

Eisend, Martin (2008), “Explaining the Impact of Scarcity Appeals Kristofferson, Kirk, Brent McFerran, Andrea C. Morales, and in Advertising: The Mediating Role of Perceptions of Darren Dahl (2017), “The Dark Side of Scarcity Promotions: Susceptibility,” Journal of Advertising, 37 (3), 33-40. How Exposure to Limited-Quantity Promotions Can Induce Hayes, Andrew F. (2012), “PROCESS: A Versatile Computational Aggression,” Journal of Consumer Research, forthcoming. Tool for Observed Variable Mediation, Moderation, and Levav, Jonathan and Rui Zhu (2009), “Seeking Freedom through Conditional Process Modeling,” white paper, www.afhayes. Variety,” Journal of Consumer Research, 36 (4), 600-10. com/public/ process2012.pdf Mukherjee, Ashesh and Seung Yun Lee (2016), “Scarcity Appeals Hong, Sung-Mook and Slavatora Faedda (1996), “Refinement of in Advertising: The Role of Expectation of Scarcity,” Journal the Hong Psychological Reactance Scale,” Educational and of Advertising, 45 (2), 256-68. Psychological Measurement, 56 (1), 173-82. Parker, Jeffrey R. and Donald R. Lehmann (2011), “When Shelf- Howard, Shu and Roger Kerin (2007), “Reference Price and Based Scarcity Impacts Consumer Preferences,” Journal of Scarcity Appeals and the Use of Multiple Influence Strategies Retailing, 87 (2), 142-55. in Retail Newspaper Advertising,” Social Influence, 2 (1), Roy, Rajat and Piyush Sharma (2015), “Scarcity Appeal in 18-28. Advertising: Exploring the Moderating Roles of Need for Inman, J. Jeffrey and Leigh McAlister (1994), “Do Coupon Uniqueness and Message Framing,” Journal of Advertising, 44 Expiration Dates Affect Consumer Behavior?”Journal of (4), 349-59. Marketing Research, 31 (3), 423-29. Simonson, Itamar (1992), “The Influence of Anticipating Regret Inman, J. Jeffrey, Anil C. Peter, and Priya Raghubir (1997), and Responsibility on Purchase Decisions,” Journal of “Framing the Deal: The Role of Restrictions in Accentuating Consumer Research, 19 (1), 105-18. Deal Value,” Journal of Consumer Research, 24 (1), 68-79. Sinha, Indrajit, Rajan Chandran, and Srini S. Srinivasan (1999), Jung, Jae Min and James J. Kellaris (2004), “Cross-National “Consumers Evaluation of Price and Promotional Restrictions: Differences in Proneness to Scarcity Effects: The Moderating A Public Policy Perspective,” Journal of Public Policy and Roles of Familiarity, Uncertainty Avoidance, and Need for Marketing, 18 (1), 37-51. Cognitive Closure,” Psychology & Marketing, 21 (9), 739-53. Van Herpen, Erica, Rik Pieters, and Marcel Zeelenberg (2009), Kivetz, Ran (2005), “Promotion Reactance: The Role of Effort- “When Demand Accelerates Demand: Trailing the Reward Congruity,” Journal of Consumer Research, 31 (4), Bandwagon,” Journal of Consumer Psychology, 19, 302-12. 725-36. How Am I Doing? Perceived Financial Well-Being, Its Potential Antecedents, and Its Relation to Overall Well-Being Richard Netemeyer, University of Virginia, USA Dee Warmath, University of Wisconsin-Madison, USA Daniel Fernandes, Catolica-Lisbon School of Business and Economics, Portugal John G. Lynch, Jr., University of Colorado-Boulder, USA

EXTENDED ABSTRACT management stress and perceived future financial security had the The effect of financial education interventions is limited (Fer- predicted effects on well-being. Finally, we found that within the nandes, Lynch, and Netemeyer 2014). Financial literacy has a very battery of controls in the model, financial literacy showed a small small effect on financial behavior. The effect of financial literacy is negative partial effect on perceived future financial security, and had even lower among low-income groups who are the ones most in need no effect on money management stress. This finding is consistent for interventions to improve their financial health. with existing literature on the lack of relationship between financial Financial matters are a problem for many consumers. The Pew knowledge and financial outcomes. Research Center (2015) reports that 55% of American households The measures we developed are reliable, valid, and brief. These can replace less than one month of their income through their sav- measures could be used by a variety of agencies and firms special- ings, and a 2016 PWC study found that 45% of employees stated that izing in savings and retirement to gauge how consumers are feeling personal financial matters cause them the most stress in their lives; about their current and future financial situations, relate these mea- more stress than their jobs, health, or relationships combined. sures to what consumers are actually doing, and then advise those Therefore, consumer researchers need to develop interventions who feel stressed or insecure about their financial future. The con- and nudges that improve financial well-being. However, very little is structs and measures developed here would allow prioritization of known about what drives financial well-being and downstream im- interventions based on the relative opportunity with the present and plications posing a problem for policy makers about how to improve future dimensions of financial well-being. financial well-being of people. The results raise the issue of whether interventions aimed at re- The lack of an accepted definition of this construct and little ducing materialism and improving self-control would reduce current systematic examination of its antecedents and consequences ham- money management stress. In addition, whether training in planning pers efforts to address financial well-being and its role in general for money, self-efficacy / ability, and willingness to take investment well-being. From a narrative qualitative procedure and five survey- risk would enhance future financial security. The answer is “possi- based studies, the objectives of this research are to: 1) define the bly”. Though materialism is a trait/value likely internalized early construct of financial well-being; 2) posit a framework of its anteced- in life (Richins 2004, 2011), there is evidence that some traits like ents and its relationship to the consequence of general well-being; 3) self-control are malleable into early adulthood (Moffitt et al. 2011). develop and validate a measure of financial well-being (Studies 1-3); Thus, school-based training / interventions designed to increase self- and 4) empirically test the framework (Studies 4 and 5). control with money may be worthwhile (Pathak, Holmes, and Zim- We had access to qualitative research from the CFPB (2015). merman 2011). Similarly, financial skill, willingness to take invest- They interviewed 59 adult consumers. Interviews were transcribed, ment risks, and planning for money are skills that can be acquired sorted, coded, and then catalogued using computer-aided qualitative (Hader et al. 2013; Perry and Morris 2005). data analysis software. Building from existing literature and this Financial education efforts in these domains may be at least as qualitative data, we conceptualize financial well-being as two sepa- effective, if not more so, than acquiring content knowledge about in- rate, but related, constructs: 1) stress related to the management of terest rates, bonds, etc. (Fernandes et al. 2014). Thus, designing and money today (current money management stress); and 2) a sense of testing programs to increase financial skill, willingness to take insecurity in one’s financial future (future financial security). vestment risks, and planning for money leading to intended changes Using three survey studies, we developed measures of these in future financial security seem to be worthwhile undertakings for constructs and demonstrated predicted correlations with other con- academics and policymakers alike (CFPB 2015). structs in the nomological net. We reduced the number of items to 10 The measures developed in the present research – future finan- (five for each construct of the scale). cial security and money management stress collectively explained With two additional studies (Study 4 and Study 5), we offered 33% (Study 4) and 39% (Study 5) of the variance in well-being. and supported a model of potential antecedents and consequences Recall that prior to our studies, we found little evidence examining of current money management stress and future financial security. the financial well-being / psychological well-being linkage. We also Importantly, we demonstrated that current money management stress observe that the effect of current money management stress on well- and future financial security are influenced by different antecedents, being is stronger for low-income individuals, and that the effect of but shared the same outcome of well-being. As such, the primary future financial security is stronger for younger individuals. There- thesis of this paper relates to the interplay of stress about money fore, specific interventions aimed at current money management for matters today and a sense of security about a financial future in pro- low-income individuals or future financial security for younger inducing feelings of financial well-being and, ultimately, emotional / dividuals are more promising at improving the well-being of these psychological well-being. individuals. Current money management stress was predicted by traits and This study represents only an initial investigation of what traits, behaviors that are negative and more short-term oriented: making behaviors, and circumstances produce higher or lower levels of well- only minimal payments, lacking in self-control, and being materi- being and the implications of financial well-being for overall well- alistic. Behaviors and traits that reflect longer-term thinking, such being. It is our intent that the concepts and findings presented here as planning for money long-term and a willingness to take invest- will instigate additional work and ultimately lead to a working model ment risk, were related to future financial security. Current money driving policy and program decisions.

Advances in Consumer Research 780 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 781 REFERENCES Perry, Vanessa. G., and Marlene D. Morris (2005), “Who is in Consumer Financial Protection Bureau (CFPB) (2015), “Financial Control?: The Role of Self-Perception, Knowledge, and Well-Being: The Goal of Financial Education,” http://www. Income in Explaining Consumer Financial Behavior,” Journal consumerfinance.gov/reports/financial-well-being/ of Consumer Affairs, 39 (2), 299-213. Fernandes, Daniel, John G. Lynch, and Richard G. Netemeyer Pew Research Center 2015. The Precarious State of Family Balance (2014), “Financial Literacy, Financial Education, and Sheets. http://www.pewtrusts.org/~/media/Assets/2015/01/ Downstream Financial Behaviors,” Management Science, 60 financialsecurityM_Balance_Sheet_Report.pdf (8), 1861-1883. PwC (2016). Employee Financial Wellness Survey. https://www. Hader, Liat, Sanjay Sood, and Craig R. Fox (2013), “Subjective pwc.com/us/en/private-company-services/publications/assets/ Knowledge in Consumer Financial Decisions,” Journal of pwc-2016-employee-wellness-survey.pdf. Marketing Research, 50 (3), 303-316. Richins, Marsha (2004), “The Material Values Scale: Measurement Moffitt, Terrie E., Louise Arseneault, Daniel Belsky, Nigel Dickson, Properties and Development of a Short Form,” Journal of Robert J. Hancox, HonaLee Harrington, Renate Houts, Consumer Research, 31 (June), 209-219. Richie Poulton, Brent W. Roberts, Stephen Ross, Malcolm R. Richins, Marsha (2011), “Materialism, Transformation Sears, W. Murray Thomson, and Avshalom Caspi (2011), “A Expectations, and Spending: Implications for Credit Use,” Gradient of Childhood Self-Control Predicts Health, Wealth, Journal of Public Policy & Marketing, 30 (2), 141-156. and Public Safety,” Proceedings of the National Academy of Sciences, 108 (7), 2693–2698. Pathak, Payal, Jamie Holmes, and Jamie Zimmerman (2011), Accelerating Financial Capability Among Youth: Nudging New Thinking. New America Foundation. http://www. newamerica.net/sites/newamerica.net/files/policydocs/ AcceleratingFinancialCapabilityamongYouth.pdf. Unconventional Consumption Methods and Enjoyment of Things Consumed: Recapturing The “First Time” Experience Ed O’Brien, University of Chicago, USA Robert Smith, Ohio State University, USA

EXTENDED ABSTRACT pothesis was three-fold: (i.) chopsticks may enhance enjoyment of Jorge Spielman’s dinner guests had an idea: to honor their host, the food; and if so, this should be (ii.) mediated by corresponding himself a blind clergyman, they dined blindfolded. In doing so they boosts in immersion akin to a revitalized “first time” experience, and found that the same foods suddenly tasted better, as if depriving their also (iii.) moderated by time, such that enjoyment and immersion sight heightened other senses. After many gatherings to similar ef- should wane once using chopsticks is no longer novel. fects, Spielman was inspired to open the world’s first “pitch black” Method restaurant in his hometown of Zürich, where patrons must eat in Sixty-eight adult participants were randomly assigned to one complete darkness. The restaurant has garnered a wealth of praise of two between-subjects conditions. The study was ostensibly about and profit since opening in 1999, and many eateries around the globe snacking speed and getting people to eat more slowly. To allegedly now imitate Spielman’s idea with similar success (Forbes 2015). achieve this goal, “traditional” participants were instructed to eat 10 Is there something special about darkness that enhances gastro- kernels of popcorn one-at-a-time using their hands. This marked a nomic pleasures? Perhaps. However, we explore a broader feature single “trial,” and they completed two trials in total. “Unconven- that might more parsimoniously explain this phenomenon: the mere tional” participants followed identical instructions but instead had to fact that dining in the dark is strange. We find that consuming an ob- use chopsticks to eat for both trials. This cover story and manipula- ject via unusual means can boost enjoyment. Beyond darkness per se, tion helped disguise the hypothesis while also accounting for various many familiar experiences might be revitalized (at least temporarily) incidental features (e.g., eating by the handful in one condition but simply by consuming them in some unconventional way. not in the other). We suggest that unconventional methods invite a “first time” After eating the popcorn in each trial, participants indicated perspective on the consumption object; even if a person has eaten their experienced enjoyment for the popcorn by rating how enjoy- a particular dish before, the overall event of eating it like this likely able, positive, and delicious it tasted, and how much they savored feels unprecedented. Accordingly, perceived novelty leads people to and liked it, each on individual scales anchored at 1 (not at all) to 9 attend more closely, engage more fully, and generally immerse them- (extremely). These dependent variables were collapsed into an enjoy- selves into an experience (Redden 2008; Sansone et al. 1992), and ment scale (α=.91). Participants also reported their immersion dur- immersion can facilitate enjoyment (Brown and Ryan 2003; Csik- ing eating by rating how much the task led them to immerse in the szentmihalyi 1990; Killingsworth and Gilbert 2010). People com- experience and focus on the food, and how much it intensified the monly lament the inability to recapture “first time” experiences with flavors, on the same scale. These mediator items were collapsed into now-familiar things, but perhaps consuming them in new ways can an immersion scale (α=.84). partly help. We investigated this possibility by asking people to eat a snack Results (Experiment 1), consume a drink (Experiment 2), and watch a video A Repeated Measures GLM with Condition (traditional, uncon- (Experiment 3) by either normal or unusual means. We hypothesized ventional) as a between-subjects factor and Time (Trial 1 and Trial 2 that, when the consumption object is familiar and the consumption rating scales) as a within-subjects factor revealed no main effect of method is new, unconventionality may revitalize enjoyment for the Condition, F(1, 66)=1.82, p=.182, η2=.03, or Time, F(1, 66)=2.16, thing consumed. Throughout, we also tested for various mediation p=.146, η2=.03; on enjoyment, but did reveal the hypothesized sig- and moderation evidence for the driving role of “first time” immer- nificant interaction between these variables, F(1, 66)=4.88, p=.031, sive feelings rather than something about individual consumption η2=.07. Pairwise comparisons reveal that unconventionality indeed methods (e.g., an unusual method may simply be an objectively enhanced enjoyment at Trial 1: the same popcorn tasted significantly superior way to consume). To the extent that the same unconven- better when eaten one-at-a-time using chopsticks (M=7.23, SD=1.08) tional method no longer produces feelings of a “first time,” the boost versus one-at-a-time using one’s hands (M=6.45, SD=1.45), F(1, should disappear (Experiments 1-2). Furthermore, unconventional 66)=6.25, p=.015, d=.61. In contrast, these boosts in enjoyment dis- methods should be especially beneficial after people have adapted appeared at Trial 2, at which point the eating experience returned to to the consumption object (Experiments 2-3). Finally, not all unusual baseline: the popcorn was equally enjoyable regardless of whether consumption methods should enhance enjoyment. Because we hy- people ate it using chopsticks (M=6.57, SD=1.63) or using hands pothesize that unconventionality may boost enjoyment by increasing (M=6.59, SD=1.24), F(1, 66)=.004, p=.950, d=.01. This suggests immersion when it is otherwise waning, unusual methods that disrupt that chopsticks enhance enjoyment due to their novelty. Other ac- immersion should not boost the experience (Experiment 3). We seek counts (e.g., if chopsticks encourage slower eating, provide effort to establish a general effect of unconventionality, highlighting when justification, or otherwise reflect an inherently superior way of eating and how unconventional consumption methods could be most help- popcorn) predict that these initial boosts should remain. Results on ful. immersion followed a similar pattern and mediated the enjoyment

boosts at Trial 1, (Effect=.53, SE=.22, 95% CIbootstrapping [.16, 1.00]; Experiment 1: Eating Popcorn with Chopsticks Hayes 2013). In Experiment 1, participants were asked to repeatedly eat Experiment 1 demonstrates that unconventional consumption popcorn using their hands (a conventional way to eat popcorn) or methods can boost enjoyment by facilitating immersion. Otherwise chopsticks (an unconventional way to eat popcorn). All of these par- identical popcorn tasted better when eaten with chopsticks than with ticipants have likely eaten popcorn before, but presumably very few hands. Moderation by time and mediation by immersion both sup- participants have eaten this familiar snack using chopsticks. Our hy- port the proposed framework: chopsticks enhance enjoyment be-

Advances in Consumer Research 782 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 783 cause they provide an unusual and therefore immersive method of for breaking hedonic adaptation when people grow to take an entity consumption and not because they are inherently superior in some for granted. Otherwise identical water remained highly enjoyable to way. As the unconventionality of chopsticks wanes, so goes their in- the extent that the way people consumed it remained unconventional. fluence on hedonic experience. A post-test revealed an affective forecasting error such that only 16% of participants predicted enjoyment Experiment 3: Watching a Video Anew to be higher for popcorn consumed with chopsticks than by hand. Experiment 3 further examined an important and intuitive boundary condition: not all unconventional methods enhance enjoy- Experiment 2: Sipping Water in Unconventional Ways ment. Participants were asked to watch the same video repeatedly. At In Experiment 2, participants were asked to generate their own the final exposure, some watched the video as they had for preceding ideas for unusual consumption methods so as to generalize the ef- exposures. Others watched this repetition in one of two “first time” fects beyond any one method concocted by an experimenter. We unconventional ways, one of which impeded immersion (watching also sought to further highlight the role of “first time” feelings in upside-down). While novel consumption methods should generally driving the boost. As proposed, unconventionality should enhance boost enjoyment by boosting immersion, we hypothesized that this enjoyment to the extent that it invites an immersive experience. Ac- form of unconventionality would not allow people to immerse into cordingly, if people generate their own lists of unusual methods, it is the experience, and so these participants may not enjoy the video unlikely that each and every idea will enhance enjoyment relative to more than normal. conventional consumption, but our hypothesis suggests that rotating Method through them (i.e., having many individual “first time” experiences Three hundred participants were recruited via Mturk. Partici- rather than relying on a single method) should maximize enjoyment pants watched an exciting video of a motorcycle ride three times, over time. This further serves our goal of testing when unconven- and rated their enjoyment after each viewing via a similar enjoy- tional methods should be most effective: rather than any one method ment scale and immersion scale as previous experiments. Before the being objectively better, we seek to highlight the critical role of “first third viewing, participants were assigned to condition. “Traditional” time” feelings. participants simply continued this process again for the third view- Method ing. “Unconventional-immersive” participants were told to watch Three hundred participants were recruited via Mturk. Partici- using “hand goggles”—forming circles with their thumbs and index pants grabbed a bottle of water and then listed five unique unconven- fingers around their eyes and using them to track the ride (e.g., to tional ways in which they had never consumed water before, with the bob their heads left or right when the driver turned left or right). restrictions that the ways could not objectively change the water it- “Unconventional-disruptive” participants also watched the video in self (e.g., no adding flavors) and that they should be able and willing a new way, but one that should inhibit immersion: the video was to actually drink using each method if instructed to do so. Sample re- flipped upside-down. sponses include “Use a funnel and drink from tap” and “Drink from Results a spoon”. Participants were then randomly assigned to condition A Repeated Measures GLM analyses with Condition as a be- and sipped and rated the water five times. “Traditional” participants tween-subjects factor and Time (View 1-3 rating scales) as a within- sipped the water in the normal way they usually drink for each of subjects factor revealed no main effect of Condition on enjoyment, the five sips. “Unconventional-variety” participants sipped the water F(2, 297)=1.04, p=.355, η2=.01, a main effect of Time such that using different methods they listed, which were piped back one at a the video grew less enjoyable across exposures, F(1, 297)=17.33, time at random. “Unconventional-repetition” participants sipped the p<.001, η2=.06, and the hypothesized interaction, F(2, 297)=9.30, water using one of the methods they listed which was piped at ran- p<.001, η2=.06. dom for the first sip and then piped again repeatedly for the remain- Pairwise comparisons reveal a significant effect at the third ing sips. After each sip, all participants rated their enjoyment for the viewing where the manipulation took place, F(2, 297)=6.39, p=.002, water on a similar scale as in Experiment 1 (α=.97). η2=.04. Unconventional participants who watched via hand goggles Results enjoyed the video significantly more (M=5.75, SD=2.21) than par- A Repeated Measures GLM with Condition (traditional, uncon- ticipants who watched the same video normally (M=4.99, SD=2.50), ventional-variety, unconventional-repetition) as a between-subjects t(297)=2.22, p=.028, d=.32. Critically, not all unconventionality factor and Time (Sip 1-5 rating scales) as a within-subjects factor did the trick: unconventionality involving watching upside-down revealed a main effect of Condition,F (2, 297)=6.68, p=.001, η2=.04, (M=4.58, SD=2.41) did not boost enjoyment versus watching nor- a main effect of Time such that the water grew less enjoyable across mally, t(297)=1.23, p=.218, d=.17. Results on immersion revealed sips, F(2, 297)=53.64, p<.001, η2=.15, and the hypothesized interac- a similar pattern, and mediated the enjoyment boosts at Trial 3, (Ef- 2 tion, F(2, 297)=11.37, p<.001, η =.07. fect=.37, SE=.13, 95% CIbootstrapping [.12, .61]; Hayes 2014). As can be seen, the “unconventional-variety” condition en- Experiment 3 replicates and extends our general framework. hanced enjoyment for all but the initial sip: for each subsequent sip, Again, boosts in enjoyment derive from boosts in immersion into participants who had many first-time experiences felt sustained en- otherwise familiar experiences, and these boosts in immersion can joyment throughout the taste test, significantly more than those who be gleaned from unusual consumption methods. But accordingly, drank normally each time (ts≥2.00, ps≤ 046, ds≥ 29) and those who consumption methods that are unconventional but disruptive for im- drank in the same unconventional way each time (ts≥2.60, ps≤.010, mersion often fail to boost enjoyment as intended. ds≥.34). Likewise, repeating the same unconventional method made the water no more enjoyable than drinking it normally for each of General Discussion these subsequent sips (ts≤.85, ps≥.395, ds≤.13). We find converging evidence across various domains of experi- These findings emphasize the driving role of unconventionality ence, direct effects, moderation analyses, mediation analyses, and in boosting enjoyment, beyond any one specific method per se. Sec- control comparisons, all highlighting the same basic effect: other- ond, these findings reveal that unusual methods are especially useful wise identical entities can become more enjoyable merely when con- 784 / Unconventional Consumption Methods and Enjoyment of Things Consumed: Recapturing The “First Time” Experience sumed in an unconventional way. A familiar food tasted better when REFERENCES people used chopsticks to eat it (Experiment 1); a familiar drink Brown, K. W., & Ryan, R. M. (2003). The benefits of being present: tasted better when people took sips in continually new ways (Experi- Mindfulness and its role in psychological well-being. Journal ment 2); and a familiar video was more enjoyable when people used of Personality and Social Psychology, 84, 822-848. hand goggles to re-watch it (Experiment 3). Importantly, however, Campbell, T., O’Brien, E., Van Boven, L., Schwarz, N., & Ubel, when unconventional methods grew dull (Experiment 1, Experiment P. A. (2014). Too much experience: A desensitization bias 2) or inhibited immersion (Experiment 3), these boosts disappeared. in emotional perspective taking. Journal of Personality and This should not have been the case if unusual methods per se were Social Psychology, 106, 272-285. objectively superior in some way, which would obviously explain Csikszentmihalyi, M. (1990). Flow: The psychology of optimal why they enhanced enjoyment. Rather, the current experiments gen- experience. New York, NY: Harper Perennial. erally suggest that such boosts may reflect rejuvenated feelings of a Forbes (2015). Dining in the dark: Basel’s Blindekuh “first time” consumption experience. Restaurant. Retrieved from: http://www.forbes.com/sites/ karlaalindahao/2015/07/30/dining-in-the-dark-at-blindekuh- Practical concerns basel-switzerland-2015/ The most important implication pertains to how these find- Hayes, A. F. (2013). Introduction to mediation, moderation, and ings bear on the psychology of waste. Waste is a growing societal conditional process analysis: A regression-based approach. concern, around the globe and in all areas of life; see recent calls New York: Guilford Press. by World Bank (2012) and the OECD (2014) for the need to better Herrnstein, R. J., & Prelec, D. (1991). Melioration: A theory of understand how to combat these increasingly wasteful trends. When distributed choice. The Journal of Economic Perspectives, 5, something declines in enjoyment, people notoriously assume that “it” 137-156. is the problem and therefore abandon the entity for a new alternative Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is (Campbell et al. 2014; Herrnstein and Prelec 1991). Clearly more rean unhappy mind. Science, 330, 932. search is needed to tackle these diverse trends, but the current studies OECD (2014). Society at a Glance 2014: OECD Social Indicators. reveal an overarching possible fix. Before intuitively (and perhaps OECD Publishing. http://dx.doi.org/10.1787/soc_glance-2014- mistakenly) reverting to a substitution strategy, simply changing the en way “it” is experienced could reduce cost and prolong value. Beyond Quoidbach, J., Mikolajczak, M., & Gross, J. J. (2015). Positive waste, the current findings more generally suggest an easy, cheap, interventions: An emotion regulation perspective. and potentially effective way to better enjoy the things we consume: Psychological Bulletin, 141, 655-693. using unconventional methods, but not overusing them, may serve to Redden, J. P. (2008). Reducing satiation: The role of categorization boost enjoyment and combat hedonic adaptation. level. Journal of Consumer Research, 34, 624-634. Sansone, C., Weir, C., Harpster, L., & Morgan, C. (1992). Once a Outstanding questions boring task, always a boring task? Interest as a self-regulatory Beyond these broader issues, the idea of unconventional con- mechanism. Journal of Personality and Social Psychology, 63, sumption raises many fruitful avenues for research. What are the 379-390. parameters of unconventionality? Experiment 3 highlights how Schumann, D. W., Petty, R. E., & Clemons, S. (1990). Predicting methods that inhibit immersion can backfire, so it would be valuable the effectiveness of different strategies of advertising variation: to identify the most important sources of inhibition. Experiment 1 A test of the repetition-variation hypotheses. Journal of suggests that some boosts could pass very quickly, so future research Consumer Research, 17, 192-202. should also identify how and why feelings of unconventionality can World Bank (2012). What a waste: A global review of solid waste be maintained. management. Washington, DC. Finally, the current findings may parsimoniously explain existing wellbeing-boosting factors, which largely appear throughout the literature as isolated strategies and effects (see Quoidbach, Mikolajczak, & Gross 2015). If such manipulations invite people to behave atypically (which they often do, see Quoidbach et al. 2015 for a review), many factors touted to boost enjoyment might largely reflect the unconventionality of the method. Until these possibilities are tested, consider using the wrong utensils at tonight’s meal rather than splurging at a trendy “dark” restaurant—it may taste just as delicious. European Consumers and their Persistent Resilience in the Face of Austerity Deirdre M. O’Loughlin, University of Limerick, Ireland Morven G. McEachern, University of Huddersfield, UK Isabelle T. Szmigin, University of Birmingham, UK Kalipso Karantinou, Athens University of Economics and Business, Greece Belem Barbosa, University of Aveiro, Portugal Maria Eugenia Fernández Moya, CUNEF, Spain Grigorio Lamprinakos, Athens University of Economics and Business, Greece

EXTENDED ABSTRACT enquiry (Denzin and Lincoln 1994; Miles and Huberman 1994) to While much has been written on austerity from political and explore the effects of austerity over a lengthy period, in terms of economic perspectives (e.g. Blyth 2013), there has been limited mar- changes to everyday consumption practices and experiences, both keting attention regarding the effects of austerity on consumers. Even prior to and during austerity. The interviews ranged in length from 45 less focus has been directed at long-term and proactive responses minutes to 2 hours and reached a deep level of richness and insight. to austerity which, we argue, can be understood via the concept of Following transcription and, where necessary, translation, cycles of persistent resilience (Golubchikov 2011). This study explores the na- coding, cross-checking and theoretical review (Strauss and Corbin ture of this enduring and dynamic form of resilience manifested by 1990; Clark and Braun 2013) were followed by all research team consumers across EU countries affected by austerity. Drawing on re- members. Using a thematic approach, three distinct themes of self- cent spatial studies, we contribute to academic literature by advanc- efficacy, creativity and community relating to persistent resilience ing our understanding of the concept of persistent resilience and by were identified. distinguishing its underlying characteristics within the consumption Persistent resilience was demonstrated through an active rene- context of EU austerity. gotiation of and continuous adaptation to everyday life challenges There is an extensive literature on resilience and its varying as well as in a process of transformation and long-term adjustment, definitions are represented from many areas including psychology often leading to a renewed sense of wellbeing, sustainable values and (e.g. Lazarus 1993), sociology (e.g. Evans and Reid 2013), consumer reciprocal goals. Self-efficacy was identified by participants’ ability research (e.g. Ball and Lamberton 2015); disaster studies (e.g. Baker to persist, adapt and transform, often facilitated through a mind-set and Baker 2014) and ecology (e.g. Adger 2000). Recent conceptu- change. Manifested through upskilling, working several jobs and en- alisations of resilience are premised not upon a one-off event from terprise formation, self-efficacy was seen as part of a process of flex- which individuals bounce back, but rather on their ability to con- ible adaptation to the risky austerity context. As part of this transfor- tinually re-emerge from on-going adversity, for which one must be mation, the restructuring of everyday practices was also undertaken continuously prepared and increasingly respond to (Neocleous 2012; through numerous creative means, reflected by the theme creativity, Evans and Reid 2013). Thus, the growing importance of a sustained, involving creative thinking, making and doing, and resourceful bud- flexible and dynamic response to the enduring challenges of modern geting to demonstrate a persistent creative response to the financial society paves the way for a more relevant form of resilience to ex- challenges of austerity. Captured by the theme community, persistent plore in our austerity context - persistent resilience. resilience was further enabled through drawing on key relationships Defined as “a form of engagement with more enduring, last- with family, friends and colleagues to share the burden of financial ing challenges and pressures…that underlie everyday life”, persis- uncertainty and, together, increase strength and resilience. Com- tent resilience allows for a proactive and continuous renegotiation munity was also reflected through deep solidarity towards others at of everyday life practices and relationships (Golubchikov 2011, 4). a localized level and also broader community and national levels. Persistent resilience was conceived within spatial studies in response Collectively, the themes of self-efficacy, creativity and community, to the pressures of modern society but also as a coping strategy over underpinning persistent resilience, reflect a wide range of proactive, prolonged periods of time. More recent studies have applied persis- continuous and sustained responses to the on-going challenges of tent resilience to explore how households and communities “develop austerity. responses to more long-term processes, such as the changing nature As European consumers continue to navigate long-term auster- of employment and … altering state policies” (Andres and Round ity, our study presents rich and timely insights into the nature of their 2015, 676). The associated restructuring of common, everyday prac- persistent resilience and how this differs across culture and historical tices and relations may, in turn, form the basis for more significant context. Reflecting the dynamic, enduring and proactive nature of transformation and even permanent change at both an individual and this resilience, our analysis identifies a wide range of self-efficacious, societal level (De Verteuil and Golubchilov 2016). Building upon creative and community-based strategies involving adapting to, per- this spatial research, we propose that the ongoing phenomenon of sisting with and transforming consumers’ everyday practices and so- austerity in affected EU states necessitates the activation of persistent cial relations. Persistent resilience in our austerity context involves resilience and we explore its specific nature and characteristics with- an enduring self-belief and positivity to creatively embrace continu- in an austerity consumption context, deliberately including in our ous change and social reciprocity at personal, local and societal lev- study countries that have experienced varying degrees of austerity. els. While the economic and social pressures of austerity may, in Our cross cultural interpretive study was conducted by a team some instances, still bring about marginalization, exclusion and dis- of European researchers across six countries. A purposive sample placement, persistent resilience provides the armor to continuously (Patton, 1990) of 38 people across a range of demographics e.g. age and creatively transform, and, thus, mitigate its effects. Regarding (ranging from 18 to 65), gender (14 male and 24 female), life-stage, addressing potential limitations of the study, the size of subgroups profession and income, including unemployed people, were inter- within the cross-cultural sample could be increased by conducting viewed from Ireland, UK, Spain, Portugal, Italy and Greece. Face to larger scale multi-country qualitative research comparing differing face in-depth interviews were chosen as an appropriate form of social levels of resilience. Future research should focus on investigating

Advances in Consumer Research 785 Volume 45, ©2017 786 / European Consumers and their Persistent Resilience in the Face of Austerity how persistent resilience could be nurtured individually and collec- Denzin, Norman K., and Yvonna S. Lincoln (1994), Handbook of tively through collaboration between governments and local com- Qualitative Research. Thousand Oaks: Sage. munity bodies. DeVerteuil, Geoff & Oleg Golubchikov (2016), “Can Resilience be Redeemed?”, City, 20 (1), 143-51. REFERENCES Evans, Brad and Julian Reid (2013), “Dangerously Exposed: Adger, Neil (2000), “Social and Ecological Resilience: Are They The Life and Death of the Resilient Subject”, Resilience, Related”, Progress in Human Geography, 24 (3), 347-64. International Policies, Practices and Discourses, 1 (2), 83-98. Andres, Lauren and John Round (2015), “The Role of ‘Persistent Golubchikov, Oleg (2011) “Persistent Resilience: Coping with Resilience’ within Everyday Life and Polity: Households the Mundane Pressures of Social or Spatial Exclusion: Coping with Marginality within the ‘Big Society’”, Introduction to a Special Session.” Paper presented at RGS- Environment and Planning A, 47, (3), 676 – 90. IBG Annual International Conference, 2 September 2011. Baker, Stacey, M. and Courtney N. Baker (2014), “Narratives of London. Cultural Trauma (and Resilience): Collective Negotiation of Lazarus, Richard. S. (1993), “From Psychological Stress to the Material Wellbeing in Disaster Recovery”, in NA - Advances Emotions, A History of Changing Outlooks”, Annual Review of in Consumer Research, 42, eds. June Cotte and Stacy Wood, Psychology, 44 (1), 1-21. Duluth, MN: Association for Consumer Research, 7-8. Miles, Matthew B. and A. Michael Huberman (1994), Qualitative Ball, Joan and Cait Lamberton (2015), “Rising Every Time Data Analysis: An Expanded Sourcebook. Thousand Oaks: They Fall: The Importance and Determinants of Consumer Sage. Resilience”, in NA- Advances in Consumer Research, 43, eds. Neocleous, Mark (2012), “Don’t Be Scared, Be Prepared” Trauma- Kristin Diehl and Carolyn Yoon, Duluth, MN :Association for Anxiety-Resilience”, Alternatives, 37 (3), 188-98. Consumer Research, 192-93. Patton, Michael Quinn (1990), Qualitative Evaluation and Blyth Mark (2013), Austerity: The History of a Dangerous Ide., Research Methods. Thousand Oaks: Sage. New York: Oxford University Press. Strauss, Anselm, and Juliet M. Corbin (1990), Basics of Qualitative Clarke, Victorial and Virginia Braun (2013) “Teaching Thematic Research: Grounded Theory Procedures and Techniques. Analysis: Overcoming Challenges and Developing Strategies Thousand Oaks: Sage. for Effective Learning”, The Psychologist, 26 (2), 120-23. Moral Consumers and the Moral Economy Mansour Omeira, University of Neuchâtel, Switzerland Valéry Bezençon, University of Neuchâtel, Switzerland

EXTENDED ABSTRACT P5: The difference between market economy and moral Moral consumers are typically understood as those individuals economy relates to the value(s) under consideration who engage in moral (ethical) consumption (Komarova Loureiro et al. 2016). In the current paper, we develop theoretical propositions P6: The struggle between market economy and moral econ- regarding moral consumers and their interface with the moral econo- omy relates to value conflict and can be investigated as my. After considering two aspects of moral persons, as moral agents such and as moral subjects, we identify a first pathway of the moral consumer – moral economy interface, from consumers to the economy. What kind of persons are moral consumers? The typical focus We then argue that moral consumers include both natural and legal is on the final consumption of individuals (natural persons). From a persons, and identify a second pathway of the interface, from the macro perspective, final consumption is undertaken by institutional economy to consumers, before concluding with research implica- units such as households or the government (European Commission tions. et al. 2009). In addition to final consumption, intermediate consump- Consumers are moral, in the sense of having moral person- tion during production can be investigated. Can we consider legal hood, independently of whether their consumption is deemed moral entities as moral persons? There has been a growing recognition that or immoral. Someone with moral personhood can be seen from two corporations have a mind, and alongside it the notion of corporate aspects: as moral agent and as moral subject (Gallagher 2007). As crime has increasingly featured in legislation (Tombs 1999). moral agents, consumers can take responsibility for their actions. As moral subjects or subjects of moral worth, they have rights and are P7: Both final and intermediate consumers are moral con- owed respect. Is every moral agent a moral subject, and is every mor- sumers al subject a moral agent? Some moral agents are not moral subjects. Humans forced into slavery conditions are moral agents who are de- nied the status of moral subjects. Some moral subjects, like biodi- P8: Moral personhood encompasses both natural and legal versity or culture, are not moral agents. Moral subjects who are also persons moral agents can respond to moral transgressions (Xie et al. 2014), such as product harm crises (Lu et al. 2016), through retaliation. We have already noted that consumers can respond to the moral transgressions of enterprises. Yet such moral transgressions can also P1: Ethical (moral) marketing involves the recognition that transform consumer moral norms. A prominent example relates to consumers are both moral agents and moral subjects the 2007 financial crisis, which came as people engaged in unethical behavior to receive bigger pay (Lewis et al. 2010). The pay structure focused on bonuses, which were unequally distributed (Crotty P2: Moral transgressions entail undermining the status of 2009). A bonus structure means that a certain pay is contingent on the one or more beings as moral subject(s) achievement of set goals; such goal-setting increases the likelihood of engaging in unethical behavior (Schweitzer et al. 2004). A combination of income and status inequality and variable pay is likely to P3: Moral agency can be exercised over oneself (e.g. ethical increase moral transgressions made for higher monetary gain. The consumption) or over others (e.g. retaliation) highly visible status rewards and non-punishment of such perceived Moral consumers are moral; they are also consumers. We can transgressions generates a social norm that challenges existing moral consider different aspects or dimensions of consumption and other norms, and may eventually lead to habit change. Our approach pro- economic activities, such as the ecological, social, political, or cul- vides a new interpretation of the finding that consumers assign less tural. Although analysis entails separating the different aspects, they blame to people who morally transgressed if the transgression was are inseparable in the real world. Assessing the morality of the activ- to gain a large enough sum of money (Xie et al. 2014). That is, the ity therefore requires considering its different aspects. For example, outcome of the value conflict within a person can be tipped by the the ecological aspect can be captured in measures such as the ecolog- changing incentive structures and related evolution in social norms ical footprint. Addressing the social aspect can involve investigating around her. the social practices associated with consumption. The cultural aspect P9: Repeated exposure to the non-punishment of perceived of consumption can address its relation to value(s) (Graeber 2001), moral transgressions modifies moral norms which are socially constructed. It enables consideration of the struggle between market economy and moral economy from a(n) (inter) subjective perspective, as a value conflict that may take the form of P10: Changes in the incentive structures of the market econo- psychological tension in the individual (Burroughs and Rindfleisch my can alter how consumers resolve their value conflicts 2002), or of political conflict in society. More importantly, it permits investigating how such conflict is managed (Miller 2008). Our investigation into what it means to be a moral consumer has broad implications for furthering the research agenda on morality P4: The morality of consumption, and of the economy more in and around markets. When are consumers not moral agents? The generally, is multidimensional delimitations are important, including for accountability purposes in cases of consumer rage (Fullerton and Punj 2004). Who (or what) is or is not to be considered a moral subject? Shedding light on the

Advances in Consumer Research 787 Volume 45, ©2017 788 / Moral Consumers and the Moral Economy matter has major implications, including about whether to eat ani- Komarova Loureiro, Yuliya, Julia Bayuk, Stefanie M. Tignor, mals (Loughnan et al. 2014), or to protect cultural heritage (Vlasic Gergana Y. Nenkov, Sara Baskentli, and Dave Webb (2016), and Turku 2016). Our analysis has also identified opportunities for “The Case for Moral Consumption: Examining and Expanding research into the morality of corporate consumers, such as how con- the Domain of Moral Behavior to Promote Individual sumers treat corporations as moral subjects. Taking corporations as and Collective Well-Being,” Journal of Public Policy & moral agents, interesting questions include moral disengagement Marketing, 35(2), 305-322. (Eriksson and Svensson 2016), risk-taking and the diffusion of re- Lewis, Victor, Kenneth D. Kay, Chandrika Kelso, and James Larson sponsibility (Wallach et al. 1964), and whistle-blowing (Dozier and (2010), “Was the 2008 Financial Crisis Caused by a Lack of Miceli 1985). Corporate Ethics?” Global Journal of Business Research, 4(2), 77-84. REFERENCES Loughnan, Steve, Brock Bastian, and Nick Haslam (2014), “The Burroughs, James E. and Aric Rindfleisch (2002), “Materialism Psychology of Eating Animals,” Current Directions in and Well-Being: A Conflicting Values Perspective,”Journal of Psychological Science, 23(2), 104-108. Consumer Research, 29(3), 348-370. Lu, Dong, Yide Liu, Hongfeng Zhang, and Ivan K. W. Lai (2016), Crotty, James (2009). “Structural Causes of the Global Financial “The Ethical Judgment and Moral Reaction to the Product- Crisis: A Critical Assessment of the ‘New Financial Harm Crisis: Theoretical Model and Empirical Research,” Architecture’,” Cambridge Journal of Economics, 33(4), 563- Sustainability, 8(7), 626. 580. Miller, Daniel (2008), “The Uses of Value,” Geoforum 39(3): 1122- Dozier, Janelle Brinker and Maria P. Miceli (1985), “Potential 1132. Predictors of Whistle-Blowing: A Prosocial Behavior Schweitzer Maurice E., Lisa Ordóñez, and Bambi Douma (2004), Perspective,” Academy of Management Review, 10(4), 823- “Goal Setting as a Motivator of Unethical Behavior,” Academy 836. of Management Journal, 47(3), 422-432. Eriksson, David and Göran Svensson (2016), “The Process of Tombs, Steve (1999), “Health and Safety Crimes: (In)visibility Responsibility, Decoupling Point, and Disengagement of and the Problems of ‘Knowing’,” in Invisible Crimes: Their Moral and Social Responsibility in Supply Chains: Empirical Victims and their Regulation, ed. Pamela Davies, Peter Findings and Prescriptive Thoughts,” Journal of Business Francis, and Victor Jupp, London: Macmillan Press, 77-104. Ethics, 134(2), 281-298. Vlasic, Mark and Helga Turku (2016), “‘Blood Antiquities’ European Commission, International Monetary Fund, United Protecting Cultural Heritage Beyond Criminalization,” Journal Nations, Organisation for Economic Co-operation and of International Criminal Justice, 14(5), 1175-1197. Development, and World Bank (2009), System of National Wallach, Michael A., Nathan Kogan, and Daryl J. Bem (1964), Accounts 2008, New York. “Diffusion of Responsibility and Level of Risk Taking in Fullerton, R. A. and G. Punj (2004), “Repercussions of Promoting Groups,” The Journal of Abnormal and Social Psychology, an Ideology of Consumption: Consumer Misbehavior,” 68(3), 263. Journal of Business Research, 57(11), 1239-1249. Xie, Wenwen, Boya Yu, Xinyue Zhou, Constantine Sedikides, and Gallagher, Shaun (2007), “Moral Agency, Self-Consciousness, and Kathleen Vohs (2014), “Money, Moral Transgressions, and Practical Wisdom,” Journal of Consciousness Studies, 14(5-6), Blame,” Journal of Consumer Psychology, 24(3), 299-306. 199-223. Graeber, David (2001), Toward an Anthropological Theory of Value: The False Coin of Our Own Dreams, New York: Palgrave. Similarity Focus and Support for Redistribution Nailya Ordabayeva, Boston College, USA Daniel Fernances, Católica-Lisbon School of Business and Economics, Portugal

EXTENDED ABSTRACT the poor’s living standards, and investing in initiatives that improve The inequality of wealth in the US has reached record high public services for the poor). levels, which has sparked a debate about the role of inequality in Study 2 tested the psychological process behind the effect. After the economic instability and the erosion of the middle class (Piketty listing the similarities or dissimilarities between the two pictures, par- 2011). But despite wide agreement that wealth inequality in the US is ticipants indicated their support for redistributive tax and redistribu- too high, support for redistributive policies has not been widespread, tive spending policies (as in Studies 1A and 1B). We then measured and there has been much debate about the effectiveness and the opti- several factors that could mediate the effect of similarity. Specifi- mal level of redistribution (Jost and Hunyady 2005). Preferences for cally, we measured the three hypothesized mediators including the redistribution stem from individuals’ distinct beliefs about the fair- perceived similarity of individuals’ dispositional inputs (motivation ness of unequal outcomes (Bénabou and Tirole 2006). Resistance to and hard work), justification of unequal outcomes (operationalized redistribution is driven, to a significant extent, by individuals’ belief through just-world beliefs), and perceived fairness of an unequal dis- that the unequal distribution of outcomes is fair because it results tribution. We also measured additional factors that could provide al- from natural differences in individuals’ dispositional inputs such as ternative explanations, including the perceived similarity of external hard work and effort (Alesina and Angeletos 2005). Whereas what circumstances (good luck, opportunity), the perceived similarity of drives support for, and resistance to, redistribution has been abun- individual outputs (income, wealth), the personal gain (self-interest) dantly researched, little work has examined the strategies that can from redistribution, subjective inequality (the perceived inequality effectively shift redistribution support (Brown-Iannuzzi et al. 2014; of the wealth distribution), and commitment to the equality principle Chow and Galak 2012). (that all individuals deserve equal rewards regardless of their contri- The present research examines a new strategy to boost redistri- butions). The mediation analyses showed that the hypothesized fac- bution support – focus on similarity (vs. dissimilarity). We build on tors, and not the alternative ones, mediated the effect of similarity prior studies which showed that prompting individuals to focus on (vs. dissimilarity) focus on redistribution support. Specifically, the similarity (vs. dissimilarity) in an unrelated task (e.g., while evaluat- serial mediation analysis confirmed that focusing on similarity (vs. ing pictures) can increase their perception of similarity of stimuli in dissimilarity) boosted individuals’ perceptions that people are similar subsequent judgments (e.g., the height of Mount Everest in relation to in their dispositional inputs (motivation and hard work), which, in a numeric anchor, Mussweiler 2001). We hypothesize that prompting turn, reduced the justification of unequal outcomes and the perceived a similarity focus in an unrelated task (e.g., evaluation of pictures) fairness of an unequal distribution, ultimately resulting in stronger will lead people to perceive greater similarity within the social envi- support for redistribution. ronment while they determine their preferred level of redistribution. Study 3 tested if the similarity effect would hold with a more Specifically, we predict that focusing on similarity (vs. dissimilarity) practical manipulation of similarity focus (rather than a theoretical will boost individuals’ perceptions of how similar people are in their manipulation involving abstract black-and-white pictures), which dispositional inputs (motivation and hard work), which will, in turn, policy makers could potentially adapt to shift public preferences weaken the justification of unequal outcomes and the perceived fair- for redistribution. US participants listed all the ways in which they ness of an unequal distribution, ultimately leading to greater support thought they were similar to other Americans (similarity focus confor redistributive policies. Four studies tested this prediction. dition) or different from other Americans (dissimilarity focus condi- Study 1A manipulated similarity focus in a picture task. US par- tion), after which they indicated their support for redistributive tax ticipants were shown two pictures used in previous research (Mark- and spending policies. The similarity effect generalized to this more man and Gentner 1997), and they were asked to list the similarities practical manipulation of similarity focus, thereby offering policy between the two pictures (similarity focus condition), dissimilarities makers one potential route (e.g., in their communication efforts) to, between the pictures (dissimilarity focus condition), or to describe at least temporarily, shift redistribution support. one picture in detail (control condition). Afterwards, participants in- Our findings propose an effective strategy to boost support dicated the extent to which they would support or oppose two redis- for redistribution. They thereby add to the emerging view that re- tributive tax policies (creating a new tax for individuals earning more distributive preferences and beliefs that underlie them may, to some than $1 million, creating a new tax for individuals earning more than extent, be malleable and changed by environmental stimuli (Brown- $5 million). To check if the similarity effect was contingent upon Iannuzzi et al. 2014; Chow and Galak 2012). Our work also furthers individuals’ political ideology, participants also completed a multi- the understanding of prior findings that redistribution support is low item scale (Nail et al. 2009) and a single-item scale (Jost 2006) of in racially fragmented environments (Alesina, Baqir and Easterly political ideology. Support for redistributive tax policies was stron- 1999). More generally, we hope that our findings will inspire future ger in the similarity condition than in the dissimilarity and control work on strategies that can change public perceptions of and prefer- conditions (there was no difference between the dissimilarity and ences for redistribution. control conditions). This effect remained significant when political ideology was added to the analysis, and it did not depend on (i.e. was REFERENCES not moderated by) ideology. Alesina, A., & Angeletos, G. (2005). Fairness and redistribution. Study 1B showed that the effect of similarity (vs. dissimilarity) American Economic Review, 95(4), 960–980. focus generalized beyond tax policies to Americans’ support for re- Alesina, A., Baqir, R., & Easterly, W. (1999). Public goods and distributive spending policies (expanding programs that improve the ethnic divisions. Quarterly Journal of Economics, 114, 1243- poor’s economic opportunities, expanding programs that improve 1284

Bénabou, R., & Tirole, J. (2006). Belief in a just world and Markman, A. B., & Gentner, D. (1997). The effects of alignability redistributive politics. Quarterly Journal of Economics, 121 on memory. Psychological Science, 5, 363-367. (2), 699-746. Mussweiler, T. (2001). “Seek and ye shall find”: Antecedents of Brown-Iannuzzi, J. L., Lundberg, K. B., Kay, A. C., & Payne, B. assimilation and contrast in social comparison. European K. (2014). Subjective status shapes political preferences. Journal of Social Psychology, 31, 499–509. Psychological Science, 26 (1), 15-26. Nail, P. R., McGregor, I., Drinkwater, A. E., Steele, G. M., & Chow, R. M., & Galak, J. (2012). The effect of inequality frames on Thompson, A. W. (2009). Threat causes liberals to think like support for redistributive tax policies. Psychological Science, conservatives. Journal of Experimental Social Psychology, 45, 23 (12), 1467-69. 901-07. Jost, J. T. (2006). The end of the end of ideology. American Piketty, T. (2011). Capital in the Twenty-First Century. Éditions du Psychologist, 61, 651-70. Seuil: Paris, France. Jost, J. T., & Hunyady, O. (2005). Antecedents and consequences of system-justifying ideologies. Current Directions in Psychological Science, 14 (5), 260-265. Customers’ Emotions in Service Failure and Recovery: A Meta-Analysis Chiara Orsingher, University of Bologna, Italy Sara Valentini, University of Bologna, Italy Alexandra Polyakova, University of Sussex, UK

EXTENDED ABSTRACT values is greater than the one expected from sampling error alone. For customers, service failure and service recovery situations First, among the three dimensions of justice, procedural justice is can be powerful triggers of strong emotional reactions. Consider the the dimension most strongly related to both negative (radj = - 0.36) following titles of reviews taken from Tripadvisor as an example: and positive emotions (radj = 0.48). Procedural justice has a stron- “Disgusted, disappointed, DON’T GO”, “Unexpected joy!”. Since ger relationship with emotions than distributive justice. On average, companies often need to deal with customer displays of emotions, all justice dimensions have higher effect sizes for positive emotions understanding what role do emotions play and what firms should do than for negative emotions, in absolute terms. This is also true for to manage emotions successfully represents important issues. The overall justice that has the strongest relationship with both negative last fifteen years of complaint handling research have witnessed a and positive emotions, although the magnitude is higher for positive rapid growth in the number of studies that have included custom- emotions. Positive emotions have a stronger impact on satisfaction ers’ emotional responses following service failure and recovery. The after recovery (radj = 0.56) than negative emotions (radj = - 0.40), al- evidence accumulated in these years suggests that emotions are re- though negative emotions have a strong negative impact on cumula- lated to cognitive reactions of customers to service recovery efforts tive satisfaction (radj = - 0.58). (i.e. perceived justice dimensions), and to relevant outcome variables With respect to the relationship between emotions and outcome such as loyalty, satisfaction, return intent, word of mouth, and com- variables, positive emotions are more strongly related to loyalty (radj plaint intentions (i.e., Lazarus 1966; Smith and Bolton 2002). De- = 0.48), than negative emotions (radj = - 0.26). By contrast, negative spite the significant number of empirical work, insights from this emotions are more powerful in shaping WOM (radj = - 0.30) than stream of research have not always been cumulative. Overall, these positive emotions (radj = 0.23). Finally, both positive and negative findings indicate that the impact of emotions may vary depending on emotions are strongly related to trust (radj = 0.49 and radj = - 0.44, the theoretical and methodological choices of the studies, the way respectively), although we retrieved few effect sizes for this relation- emotions are measured (Richins 1997), and the cultural orientation. ship. This suggests the need for a meta-analysis to integrate the evidence Regarding homogeneity analysis results are mixed. Although of accumulated empirical research. More specifically, through meta- only five chi-square tests are significant, all credibility intervals are analysis we aim to: 1) reflect on the theoretical conceptualization sufficiently wide (exceeding .11) or include zero (Sagie andKo- of emotions in service recovery domain, 2) map the constructs that slowsky 1993). These results suggest the presence of possible mod- have been examined in relation to emotions, 3) identify which of erator variables. these constructs are more strongly related to emotions, and 4) as- The results of the moderator analyses show that, on average, sess the role of moderating variables in shaping the magnitude of the dimensional models have higher effects sizes between emotions and relationships. their correlates. Scenarios produce on average higher effects size be- In the current research we present a meta-analysis of 331 pair- tween emotions and their correlates, whereas a large number of items wise relationships coming from 69 independent studies of the cor- generates on average lower effect sizes than a smaller number. The relates of emotions in service failure and recovery. type of participant (students vs. non-students) does not seem to affect Our results show that for negative emotions, half of the studies the size of the correlations involving emotions and the other con- relies on a discrete model (50.6%), and the remaining half (49.4%) on structs. Among the cultural values, only uncertainty avoidance and a dimensional model (Barret 1998).Within the discrete model, anger long-term orientation significantly moderate the relationships with is the emotion that is most frequently taken into consideration (52%), emotion correlates. followed by frustration, regret, and helplessness. Interestingly, for Thus, we draw four main conclusions: first, our results show positive emotions, most studies are anchored to a dimensional view that the relationships between emotions and their correlates are model (92.7%). Moreover, we found that discrete and dimensional higher when researchers use dimensional models rather than discrete constructs are measured differently. Discrete models tend to measure models of emotions. emotions using multi-item scales were each item represents different Second, the analysis of pairwise relationships highlights that nuances of the same category of emotions. Dimensional models use the perceived fairness of procedures is more powerful than tangible multi-item scales where each item represents a specific type of emo- compensation and interpersonal treatment in triggering both negative tion from a different emotion category. and positive emotions. Third, satisfaction after recovery and loyalty Additionally, negative emotion scales show a wide disper- are more strongly related to positive than to negative emotions. Fi- sion of items expressing different types of emotions across studies, nally, the findings of the moderator analysis indicate that scenarios whereas this variation is less pronounced for positive emotion scales. might inflate the relationships between emotional reactions and their For example, we found 25 different negative emotions mentions in correlates and that multi vs. single item scale lower on average the the studies vs. only 13 positive ones. Anger is by far the emotion correlations. Researchers should also note that uncertainty avoid- most frequently included (18.7%) in dimensional models, followed ing cultures generate higher effect sizes in the relationship between by disappointment and annoyance (both 7.8%, respectively). The emotions and their correlates, and suggests, for these cultures, an most common positive emotions are happiness (20.6%), followed by important role of the contact personnel and the company’s policies in joy (12.7%) and pleasure (11.8%). handling the emotional reactions of customers. We also analyzed the pairwise relationships and the average effect sizes of emotions’ correlates and the results of the homogeneity analysis to test whether the observed variation in effect size

Advances in Consumer Research 791 Volume 45, ©2017 792 / Customers’ Emotions in Service Failure and Recovery: A Meta-Analysis REFERENCES Sagie, A., & Koslowsky, M. (1993). Detecting moderators with Barrett, L. F. (1998). Discrete emotions or dimensions? The role of meta‐analysis: an evaluation and comparison of techniques. valence focus and arousal focus. Cognition & Emotion, 12(4), Personnel Psychology, 46(3), 629-640. 579-599. Smith, A. K., & Bolton, R. N. (2002). The effect of customers’ Lazarus, R. S. (1966). Psychological stress and the coping process. emotional responses to service failures on their recovery New York: McGraw-Hill. effort evaluations and satisfaction judgments.Journal of the Richins, M. L. (1997). Measuring emotions in the consumption Academy of Marketing Science, 30(1), 5-23. experience. Journal of Consumer Research, 24(2), 127-146. Grammatical Subject, Base Rates, and Persuasion Max Ostinelli, Winthrop University, USA David Luna, Baruch College, USA

EXTENDED ABSTRACT blender A,” whereas in the product-subject condition, participants Companies often make claims about the success rates of their were informed that “blender B performs much worse than blender A products and/or the failure rates of competitors’ products. We ex- for 50% of customers.” The dependent variable was the choice be- amine whether changes in the grammatical subject used in these tween the two blenders. Results revealed a significant interaction ef- claims affect persuasion. Specifically, we differentiate between user- fect. For success claims (e.g., Blender A performs better than Blender subjects, where users are the grammatical subject of the sentence, B),more participants chose Blender A in the user-subject condition and product subjects, where the product is the grammatical subject than in the product-subject condition, hence showing that the user- of the sentence. For example, the success of a weight loss program subject was more persuasive. We explain this finding by suggesting can be presented either with a user-subject claim stating that “80% of that the optimism bias cued by the user-subject makes people believe participants succeed in reducing their excess weight” (Digitale 2010) they can do better than the average consumer, thus making them or with a product-subject claim stating that “our program succeeds more likely to choose blender A. For failure claims (i.e., Blender B in reducing excess weight for 80% of participants.” Similarly, the performs worse than Blender A), more participants chose Blender A failure of competitors’ sales training program can be presented with in the product-subject condition than in the user-subject condition. a user-subject claim stating that “75% of all participants in traditional Because the claim discredited the performance of blender B, this sales training fail to learn anything” (Miller and Sinkovitz 2012) or finding suggests that the user-subject was less persuasive than the with a product-subject claim stating that “Traditional sales training product-subject (i.e., fewer [more] people chose Blender A[B]). We fails to teach anything for 75% of all participants.” suggest that optimism bias cued by the user-subject makes people We empirically demonstrate that changes in the grammatical believe they are less likely to fail than the average consumer, thus subject affect persuasion through a mediation chain that includes at- resulting in an underestimation of possible failure for blender B. tribution of causality and optimism bias as mediators. Drawing on Experiment 2 provides evidence for the underlying mecha- the linguistic literature (Fausey and Boroditsky 2010; McGlone et al. nism by testing the three-path mediation process (subject→causal 2013), we suggest that user-subject claims lead to the causal attribu- attribution→optimism bias→attitudes) using the bootstrap proce- tion of success/failure to the users of the product rather than to the dure proposed by Hayes (2012, model 6). Experiment 3 replicates product’s characteristics. the subject effect with different base rates ranging from 15% to 85%. Based on research on unrealistic optimism (Weinstein 1987), Finally, a field study conducted with Google AdWords shows that we predict that this difference in causal attribution affects peoples’ a success claim for an existing company resulted in a higher click- expectations about product performance. That is, situations per- through-rate when presented with a user-subject, than with a prod- ceived as more controllable can increase people’s expectation of out- uct-subject. performing others (Helweg-Larsen and Shepperd 2001; Weinstein This work makes several contributions. First, it contributes to 1987). Thus, user-subjects, which lead to causal attributions that are research on linguistic framing (Patrick and Hagtvedt 2012) in a sig- more controllable (i.e., performance depends on customers’ actions) nificant manner: Whereas most examples of verbal framing examine should prompt more optimism, which results in the overestimation of changes in specific words (e.g., 97% lean meat vs. 3% fat meat), our success (e.g., I will be among those who succeed) and in the underes- work examines changes in syntax, which can, in principle, be applied timation of failure (e.g., I will not be among those who fail). to a broader variety of marketing communications of product success Following these predictions, the persuasiveness of different and failure. Second, it contributes to research on causality. Previous grammatical subjects should depend on whether claims are about work on causality suggests that product attributions are more persua- success or failure. For success claims (e.g., 70% success rate), user- sive because perceived to be more stable and thus more diagnostic of subject claims should be more persuasive (e.g., result in higher at- the product future performance (Folkes 1984). We show that under titudes toward the product described) than product-subject claims, some conditions product attribution can be less persuasive than cus- since greater optimism bias increases the perceived likelihood of per- tomer attribution, hence enriching our understanding of the effects of sonal success. For failure claims (e.g. 70% failure rate), user-subject causal attribution on persuasion. claims should be less persuasive, resulting in less damaging attitudes toward the product described, since greater optimism bias decreases REFERENCES the perceived likelihood of personal failure. Digitale, Erin (2010), “$12.7 Million Grant for Study of Childhood We tested this prediction in four experiments. In Experiment 1, Obesity Treatment,” (accessed June, 20th 2015). participants read a claim about the performance of two blenders (i.e., Fausey, Caitlin M. and Lera Boroditsky (2010), “Subtle Linguistic blender A and blender B) which manipulated both the type of claim Cues Influence Perceived Blame and Financial Liability,” and its grammatical subject, resulting in a 2(claim type: success vs. Psychonomic Bulletin & Review, 17 (5), 644-50. failure) x 2(subject: user vs. product) between-subject design. Suc- Folkes, Valerie S. (1984), “Consumer Reactions to Product Failure: cess claims reported that blender A performed better than blender B. An Attributional Approach,” Journal of Consumer Research, In the user-subject condition, participants were informed that “50% 10 (March), 398-409. of customers perform much better with blender A than blender B,” Hayes, Andrew F. (2012), “ PROCESS: A Versatile Computational whereas, in the product-subject condition, participants were informed Tool for Observed Variable Mediation, Moderation, and that “blender A performs much better than blender B for 50% of cus- Conditional Process Modeling “ in Retrieved from http://www. tomers.” Failure claims reported that Blender B performs worse than afhayes.com/public/process2012.pdf. Blender A. In the user-subject condition, participants were informed that “50% of customers perform much worse with blender B than

Helweg-Larsen, Marie and James A Shepperd (2001), “Do Miller, Marc and Jason Sinkovitz (2012), Selling Is Dead: Moving Moderators of the Optimistic Bias Affect Personal or Target Beyond Traditional Sales Roles and Practices To Revitalize Risk Estimates? A Review of The Literature,” Personality and Growth: John Wiley & Sons. Social Psychology Review, 5 (1), 74-95. Weinstein, Neil D. (1987), “Unrealistic Optimism About McGlone, Matthew S., Robert A. Bell, Sarah T. Zaitchik, and Susceptibility to Health Problems: Conclusions From a Joseph McGlynn III (2013), “Don’t Let the Flu Catch Community-Wide Sample,” Journal of Behavioral Medicine, You: Agency Assignment in Printed Educational Materials 10 (5), 481-500. About The H1N1 Influenza Virus,” Journal of Health Communication, 18 (6), 740-56. The Drain of Affective Decisions Ashley Otto, Baylor University, USA Joshua Clarkson, University of Cincinnati, USA Ryan Rahinel, University of Cincinnati, USA

EXTENDED ABSTRACT of resource drain in eliciting delay related to affect- and cognitive- A fundamental premise in decision making is that decisions based decisions. based on affect are fast and economical relative to decisions based on cognition. That is, “going with one’s gut” reliably leads to quicker Experiment 1 decision making and requires less processing resources to make a Using a real-choice paradigm, Experiment 1 sought to test the decision than does a more cognitive focus such as “thinking things hypothesis that affectively-based decisions elicit greater choice de- through” (Epstein 1990). This pattern is evident from the robust find- lay. ing that constraints on processing resources such as cognitive load, Method and Results time pressure, and distraction increase the reliance on affect in deci- Ninety undergraduates (51% Male; Mage = 20.47) were recruited sion making (e.g., Nowlis and Shiv 2005; Pham, Cohen, Pracejus, to complete a study on print advertisements. Following the study and Hughes 2001; Shiv and Fedorikhin 1999). Thus, relative to its welcome, all participants were told that a consumer packaged goods cognitive counterpart, decisions based on affect are highly resource- company, whose name was ostensibly concealed for privacy pur- efficient (see Pham 2007). poses, was seeking feedback on print advertisements. Participants Despite this robust effect, we ask in this paper whether consum- were randomly assigned to receive one of three granola bar adver- ers’ perceptions of the resource efficiency of affect-based decision tisements. Importantly, across all conditions the image on the adver- match reality. In particular, we propose the possibility that consum- tisement remained the same. However, to manipulate the basis of the ers’ perceive affect- (v. cognitive-) based decisions to require more— decision, we altered the slogan printed on the ad. This manipulation rather the less—resources to arrive at a decision. This proposition is is consistent with prior research that has manipulated the affective or based on research showing that individuals tend to overestimate the cognitive basis of a decision by altering an advertisement’s slogan impact of future affect (Gilbert et al. 1998; Lowenstein and Schkade (see Cian, Krishna, and Schwarz 2015). 1999; Wilson et al. 2004; Wilson and Gilbert 2003). That is, though Importantly, after viewing the ad, all participants were told that individuals accurately predict the valence of their affect and even as a thank you for their time, they would be receiving a granola bar. the specific emotions that they will experience (i.e., joy vs. anger; Participants were asked what they would like to do—make a granola Robinson and Clore 2001), they are not able to accurately predict the bar choice now or make a granola bar choice later. For those who duration and impact of their future affective states, with the specific opted to make a choice now, they were presented with the granola tendency to overestimate these errors (e.g., Gilbert, Driver-Linn, and bar options on a subsequent screen. For those who opted to make Wilson 2002; Gilbert, Lieberman, Morewedge, and Wilson 2004; a choice later, they were presented with the granola bar options at Buechel, Zhang, and Morewedge 2014; see also Loewinstein 2005). the end of the study. Importantly, then, only the timing of the deci- Just as individuals tend to make overestimation errors regarding the sion varied, as it was clear to participants that the decision would be impact of affective responses (see Wilson and Gilbert 2003), we ar- made. gue for an overestimation error regarding the resource demand as- The choice data were analyzed via a chi-square test to analyze sociated with affectively-based decisions. potential differences in delay (0 = make choice, 1 = delay choice) as This misperception is critical because, if true, it should directly a function of the basis of the decision (0 = control, 1 = cognitive, 2 = undermine consumers’ desire to engage in the decision making pro- affective). As expected, the analysis revealed a significant difference cess. In other words, if consumers perceive an insufficiency or drain in delay based on the basis of the decision (2 (2, N = 90) = 7.78, p = of resources when making affect-based decisions, then they should .020). Follow up analyses revealed that those in the affectively-based consequently demonstrate a greater propensity to delay these deci- condition were much more likely to delay choice (60%) compared sions. Importantly, here, choice delay serves not only as an outcome to those in the cognitively-based (26.67% [2 (1, N = 60) = 6.79, p = measure but as a specific choice context that should be most sensi- .009]) or control (33.33% [2 (1, N = 60) = 4.29, p = .038]) conditions, tive to subtle differences in perceptions—if in fact perceptions do not which did not differ from one another p( > .57). align with reality. Essentially, the option to engage in choice delay allows for the opportunity for perceptions to override the general prop- Discussion erties of affectively-based decisions, which appear most apparent The findings of Experiment 1 offer initial evidence thatcon- when choice is forced (i.e., absent a no-choice option; see Shiv and sumers are more likely to delay affectively-based decisions—here, Fedorikhin 1999). Given that consumers are rarely forced to make in an actual consumption experience. Indeed, across conditions, choices, this research therefore raises the possibility that marketing participants responded to the same choice. However, those in the tactics emphasizing affect (e.g., advertisements, slogans, storylines) affectively-based condition were significantly more likely to delay could unintentionally inhibit consumers’ motivation to purchase. the choice than were those in the cognitively-based and control con- We tested this possibility in three experiments. Importantly, in ditions. This latter effect is especially interesting as it provides initial each experiment, we controlled for decision importance and difficul- evidence that an affective-basis increases consumers’ likelihood to ty, as both variables have been previously shown to impact consum- engage in delay. ers’ desire to postpone decisions (Dhar 1997; Krijnen, Zeelenberg, and Breugelmans 2015; Greenleaf and Lehman 1995). Relatedly, we Experiment 2 also rule out alternative explanations related to cognitive laziness The goal of Experiment 2 was to directly test the possibility and mood. These variables have no impact on the results, which only that consumers are miscalibrated with regard to their perceptions of further strengthen our confidence in role of consumers’ perceptions affectively-based decisions. That is, while research has repeatedly

Advances in Consumer Research 795 Volume 45, ©2017 796 / The Drain of Affective Decisions demonstrated that affectively-based decisions are less resource de- Method and Results manding (see Pham 2007), we contend that consumers’ perceive One hundred participants (51% Male; Mage = 37.23) were re- affectively-based decisions to be more (not less) resource demand- cruited through Mechanical Turk to complete a study on decision ing. Though inconsistent with reality, this prediction is consistent making. Following an introduction to the study, participants were with work demonstrating that individuals tend to overestimate the told that the study was about coffee. As in Experiment 2, participants impact and duration of future affective states (Lowenstein and Sch- were assigned to one of three decision bases conditions: affective, kade 1999; Wilson and Gilbert 2003). cognitive, or control. Specifically, those in the affectively-based con- Method and Results dition were asked to describe their feelings about purchasing coffee, while those in the cognitively-based condition were asked to describe One hundred and forty (56% Female; M = 36.14) participants age their thoughts about purchasing coffee on a series of items that were were recruited through Mechanical Turk to complete a study on deci- intentionally biased to promote agreement with these statements. sion making. All participants were told that the study was about tea Importantly, however, participants responded to each of the items and were randomly assigned to one of three conditions: affective, on 5-point scales anchored at 1 – Somewhat agree to 5 – Completely cognitive, or control. Participants in the affective and cognitive con- agree, which were intentionally biased to promote agreement with ditions were instructed to focus on either their feelings (affectively- the statements (Clarkson, Janiszewski, and Cinelli 2013; Salancik based condition) or thoughts (cognitively-based condition) in rela- 1974; Salancik and Conway 1975; Tormala and DeSensi 2008; see tion to tea (adapted from Mikles et al. 2010, 2011), while those in the Petrocelli, Martin, and Li 2010). Those in the control condition did control condition were given no further instructions. not receive a focus manipulation and were automatically forwarded Participants were then asked to imagine that they were shop- to the next portion of the study. ping at a grocery store and that they came upon the tea aisle and All participants were then asked to imagine that they were shop- noticed a few options available for purchase. Prior to making a pur- ping for coffee and noticed several options available. Participants chasing decision, participants responded to a series of questions re- were then instructed to think about the scenario and asked to make a garding the tea scenario that assessed the perceived resource drain choice between deciding on whether or not to purchase coffeenow or associated with making a decision (α = .71; e.g., How much do you whether or not to purchase coffeelater . Finally, following a series of anticipate feeling exhausted of your mental resources after consider- filler questions, participants completed the Implicit Theories about ing this decision?, How mentally exhausted do you anticipate feeling Willpower Scale (Job et al. 2010) to index participants’ willpower after considering this decision?, How much do you anticipate feeling theories (α = .85; e.g., Your mental stamina fuels itself, Even after drained of your mental resources after considering this decision?). strenuous mental exertion, you can continue doing more of it). Following these questions, participants were asked to make a choice The choice data (0 = make choice, 1 = delay choice) were sub- to decide whether or not to purchase tea now or decide whether or mitted to a hierarchical logistic regression, with the basis of the deci- not to purchase tea later. sion (0 = control, 1 = cognitive, 2 = affective) and willpower theory Participants’ choice was coded as 0 = make choice or 1 = delay (continuous, mean-centered) as main effect predictors in the first step choice. These data were then submitted to a chi-square test to com- and their interaction in the second step (Cohen et al. 2003). Repli- pare the difference in delay as a function of the basis of the decision cating the prior two studies, the analysis revealed a main effect of (0 = control, 1 = cognitive, and 2 = affective). The analysis revealed decision basis (β = .84, Wald’s 2 = 7.30, p = .007) and, as expected, a significant difference in choice delay based on the basis (2 (2, N no main effect of willpower theory (p > .98). Importantly, the results = 140) = 15.28, p < .001). Those in the affectively-based condition revealed a significant decision basis × willpower theory interaction (28.57%) were much more likely to delay choice than those in the (β = -.93, Wald’s 2 = 5.26, p = .022; see Figure 1). Consistent with ex- cognitively-based (6.12%; [2 (1, N = 91) = 8.28, p = .004]) or control pectations, limited theorists (+1 SD) were significantly more likely (4.08%; [2 (1, N = 91) = 10.42, p = .001]) conditions, which did not to delay choice in the affectively-based condition, compared to both differ from each other (p > .64). Importantly, meditational analysis the cognitively-based (β = 1.85, Wald’s 2 = 6.05, p = .014) or control confirmed a significant mediating pathway through the perceived re- (β = 1.70, Wald’s 2 = 8.56, p = .003) conditions, which did not differ source drain index (95% CI: .027, .72). from each other (p > .30). For unlimited theorists (-1 SD), there was Discussion no difference in choice delay based on the basis of the decision (ps The findings of Experiment 2 demonstrate that consumers: (i) > .83). do in fact perceive affectively-based decisions as more draining of their mental resources than cognitively-based decisions, and (ii) this Discussion The findings of Experiment 3 offer an alternative means by misperception dictated the desire to engage in the decision-making which to test the robustness of the proposed misperception associ- process. ated with affectively-based decisions. Here, we examined the role Experiment 3 of willpower theories and found that the effect of an affective ba- Experiment 2 demonstrated that consumers delay affectively- sis on choice delay was bound to those who naturally perceive they based decisions because they perceive them as being more drain- have a limited (vs. limited) amount of resources. That is, consumers ing of mental resources. Importantly, if the findings of Experiment delayed affectively-based decisions when they implicitly believed 2 hold true, then this effect should only occur for consumers who they had an insufficiency of resources (i.e., limited theorists). Impor- naturally believe their resources can be drained—that is, those who tantly, these findings only corroborate those of Experiment 2, which believe that they have a limited (vs. unlimited) storehouse of avail- demonstrated that perceptions of resource drain underlie the effect of able resources (Mukhopadhyay and Gohar 2005; Job, Dweck, and affect on choice delay. Together, then, the findings of Experiments Walton 2010). Thus, in Experiment 3, we offer an alternative test of 2 and 3 offer converging means by which to demonstrate a resource- the mechanism by examining the role of willpower beliefs (i.e., the based account for the delay of decisions based on affect. extent to which individuals believe their ability to regulate behavior is a limited or unlimited resource; see Job et al. 2010). Advances in Consumer Research (Volume 45) / 797 Conclusion Gilbert, Daniel T., Elizabeth C. Pinel, Timothy D. Wilson, Stephen Though research has reliability demonstrated that affective (vs. J. Blumberg, and Thalia P. Wheatley (1998), “Immune cognitive) based decisions require less resources to arrive at a deci- Neglect: A Source of Durability Bias in Affective Forecasting,” sion, we find that consumers are miscalibrated with regard to the per- Journal of Personality and Social Psychology, 75, 617-638. ceived resource drain of affect-based decisions. Three experiments Greenleaf, Eric A., and Donald R. Lehmann (1995), “Reasons for demonstrated that this perceived insufficiency of mental resources Substantial Delay in Consumer Decision Making,” Journal of altered consumers’ desire to engage in the decision-making process. Consumer Research, 22, 186-199. That is, perceiving a lack of sufficient resources naturally altered a Job, Veronika, Carol S. Dweck, and Gregory M. Walton (2010), consumer’s desire to make decisions. Moreover, this effect occurred “Ego Depletion—Is it all in your Head? Implicit Theories despite controlling for decision difficulty and decision importance about Willpower Affect Self-Regulation,”Psychological and independent of cognitive laziness or mood. Finally, we dem- Science, 21, 1686-1693. onstrated the robustness of this misperception by showing that the Krijnen, Job MT, Marcel Zeelenberg, and Seger M. Breugelmans effects only occurred for those who perceive they have a limited (vs. (2015), “Decision Importance as a Cue for Deferral,” unlimited) amount of resources. Judgment and Decision Making, 10, 407-415. Collectively, we believe this work offers critical insight for Loewenstein, George (2005), “Hot-Cold Empathy Gaps and marketers who all too often seek to invoke an affective-basis in con- Medical Decision Making,” Health Psychology, 24, S49. sumers’ decisions. That is, marketers alter images, storylines, and Loewenstein, George, and David Schkade (1999), “Wouldn’t slogans to elicit greater affect. Yet the present findings suggest these it be Nice? Predicting Future Feelings,” In E. Diener, strategies propel consumers to disengage from the decision-making D. Kahneman, & N. Schwarz (Eds.), Well-Being: The process. Indeed, Experiment 1 demonstrated that even subtle ad- Foundations of Hedonic Psychology, New York: Russell Sage vertisement slogans that manipulate decision bases (e.g., “To refuel Foundation Press, 85-105. your dreams” vs. “To refuel your mind”) altered whether or not con- Should I go with my Gut? Investigating the Benefits of Emotion- sumers wanted to engage in actual choice. Furthermore, by studying Focused Decision Making (2011), “Should I go with my the effects of this misperception on choice delay (vs. forced choice), Gut? Investigating the Benefits of Emotion-Focused Decision we have not only allowed for the opportunity for perceptions to over- Making,” Emotion, 11, 743-753. ride reality, but also allowed for a more naturalistic test of consumer Mikels, Joseph A., Corinna E. Löckenhoff, Sam J. Maglio, Laura decision making. L. Carstensen, Mary K. Goldstein, and Alan Garber (2010), “Following your Heart or your Head: Focusing on Emotions REFERENCES Versus Information Differentially Influences the Decisions Buechel, Eva C., Jiao Zhang, Carey K. Morewedge, and Joachim of Younger and Older Adults,” Journal of Experimental Vosgerau (2014), “More Intense Experiences, Less Intense Psychology: Applied, 16, 87-95. Forecasts: Why People Overweight Probability Specifications Mikels, Joseph A., Sam J. Maglio, Andrew E. Reed, and Lee J. in Affective Forecasts,”Journal of Personality and Social Kaplowitz (2011), “Should I Go With My Gut? Investigating Psychology, 106, 20-36. the Benefits of Emotion-Focused Decision Making,”Emotion, Cian, Luca, Aradhna Krishna, and Norbert Schwarz (2015), 11, 743-753. “Positioning Rationality and Emotion: Rationality is Up and Mukhopadhyay, Anirban, and Gita Venkataramani Johar, (2005), Emotion is Down,” Journal of Consumer Research, 42, 632- “Where There is a Will, is There a Way? Effects of Lay 651. Theories of Self-Control on Setting and Keeping Resolutions,” Clarkson, Joshua J., Matthew J. Valente, Christopher Leone, and Journal of Consumer Research, 31, 779-786. Zakary L. Tormala (2013), “Motivated Reflection on Attitude- Nowlis, Stephen M., and Baba Shiv (2005), “The Influence of Inconsistent Information: An Exploration of the Role of Fear Consumer Distractions on the Effectiveness of Food-Sampling of Invalidity in Self-Persuasion,” Personality and Social Programs,” Journal of Marketing Research, 42, 157-168. Psychology Bulletin, 39, 1559-1570. Petrocelli, John V., Jacob L. Martin, and Winston Y. Li (2010), Cohen, Jacob, Patricia Cohen, Stephen G. West, and Leona S. “Shaping Behavior Through Malleable Self-Perceptions: A Aiken (2003), Applied Multiple Regression/Correlation Test of the Forced-Agreement Scaling Effect (FASE),” Journal Analysis for the Behavioral Sciences, Mahwah, NJ: Erlbaum. of Research in Personality, 44, 213-221. Dhar, Ravi (1997), “Consumer Preference for a No-Choice Pham, Michel Tuan (2007), “Emotion and Rationality: A Critical Option,” Journal of Consumer Research, 24, 215-231. Review and Interpretation of Empirical Evidence,” Review of Epstein, Seymour (1990). Cognitive-Experiential Self-Theory. General Psychology, 11, 155-178. In L. A. Pervin (Ed.), Handbook of Personality Theory and Pham, Michel Tuan, Joel B. Cohen, John W. Pracejus, and G. Research, New York: Guilford Press, 165-192. David Hughes (2001), “Affect Monitoring and the Primacy Gilbert, Daniel T., Erin Driver-Linn, and Timothy D. Wilson of Feelings in Judgment,” Journal of Consumer Research, (2002), “The Trouble with Vronsky: Impact Bias in the 28, 67-88.Robinson, Michael D., and Gerald L. Clore (2001), Forecasting of Future Affective States,” In L. F. Barrett and “Simulation, Scenarios, and Emotional Appraisal: Testing the P. Salovey (Eds.), The Wisdom in Feeling: Psychological Convergence of Real and Imagined Reactions to Emotional Processes in Emotional Intelligence, New York: Guilford Stimuli,” Personality and Social Psychology Bulletin, 27, Press, 114-143. 1520-1532. Gilbert, Daniel T., Matthew D. Lieberman, Carey K. Morewedge, Salancik, Gerald R. (1974), “Inference of One’s Attitude from and Timothy D. Wilson (2004), “The Peculiar Longevity of Behavior Recalled Under Linguistically Manipulated Things Not so Bad,” Psychological Science, 15, 14-19. Cognitive Sets,” Journal of Experimental Social Psychology, 10, 415-427. 798 / The Drain of Affective Decisions

Salancik, Gerald R., and Mary Conway (1975), “Attitude Wilson, Timothy D., and Daniel T. Gilbert (2003), “Affective Inferences from Salient and Relevant Cognitive Content About Forecasting,” Advances in Experimental Social Psychology, Behavior,” Journal of Personality and Social Psychology, 32, 35, 345-411. 829-840. Wilson, Timothy D., Thalia P. Wheatley, Jaime L. Kurtz, Elizabeth Shiv, Baba, and Alexander Fedorikhin (1999), “Heart and Mind in W. Dunn, and Daniel T. Gilbert (2004), “When to Fire: Conflict: The Interplay of Affect and Cognition in Consumer Anticipatory Versus Postevent Reconstrual of Uncontrollable Decision Making,” Journal of Consumer Research, 26, 278- Events,” Personality and Social Psychology Bulletin, 30, 340- 292. 351. Tormala, Zakary L., and Victoria L. DeSensi (2008), “The Perceived Informational Basis of Attitudes: Implications for Subjective Ambivalence,” Personality and Social Psychology Bulletin, 34, 275-287. The Role of Scales on Evaluations of Identical Goal Progress Timucin Ozcan, Rollins College, USA Kunter Gunasti, Washington State University, USA

EXTENDED ABSTRACT ness of product attributes (Ozcan and Sheinin 2012, 2013) and have People often use different measurement scales due to personal, been shown to impact consumer evaluations and decisions. Thus, the social, cultural, economical, or technical reasons as well as simple scaleocity effect will be mediated by the completeness perceptions convenience. The naive act of using different ways to measure prog- induced by round numbers on corresponding scales. Formally, we ress may lead two people to feel very differently about their lev- posit the following hypotheses: els of goal accomplishment even when they have totally identical achievements. For example, imagine that Mark and Jessica run a Hypothesis 1: Scaleocity effect - When measuring and evalu- lap around town covering 2.23 miles of distance every day. Mark is ating identical goal progress using different keeping track of the total distance he runs in miles, whereas Jessica scales, the scale that corresponds to round nu- is counting the number of days she runs. Both of them run for 9 days merical values will lead to: i) a higher sense of completing 20 miles each. Will they feel the same way about their accomplishment, ii) decreased motivation/per- progress and willingness to continue their exercise? We propose that sistence to keep working toward a goal, and iii) Mark will have a higher sense of accomplishment as he just reached lower risk taking behavior. the 20-mile benchmark. Despite achieving identical progress, Jessica Hypothesis 2: Identical or even lower progress expressed in will not share the same sense of achievement with Mark. As Jessica lower but round numbers on a given scale might is measuring his progress differently (in days), she feels the urge to be perceived as better than equal or even high- run one more day to reach the 10-day benchmark based on her choice er progress expressed in higher but non-round of measuring progress. Further imagine that both run for another day numbers on another scale and they both complete 10 days and 22.3 miles of running. This time Jessica will be feeling great about his progress because she com- Hypothesis 3: The scaleocity effect (scale-induced perception pleted a 10-day course of running. On the other hand, Mark will not of accomplishment) will be mediated by an ar- feel very accomplished and he might feel a need to run just another tificial sense of completeness conveyed by round day to reach 25 miles and feel a sense of completeness based on his numbers measured on a scale. arbitrary choice of measurement scale. Vast literature in numerosity predicts that a more numerous We conducted five experimental studies to test our hypotheses. scale leads to perceptions of larger numerical magnitudes compared While the first three studies confirm H1 and H2, Studies 4 and 5 fur- to a less numerous scale (e.g., Burson, Larrick and Lynch 2009; Pan- ther test the mediation effect of completeness. Study 4 demonstrates delaere, Briers and Lembregts 2011). For example, 365 days might the mediation. We found that people might decide that they have not be seen as longer than 12 months because the former is more numer- performed well in a task and just to feel a sense of accomplishment ous (Monga and Bagchi 2012). Thus, 20 miles might be perceived they will work unnecessarily harder to lose additional amounts of as better progress than 9 days. On the other hand, in the case of 10 weight; run additional distances/days; spend more time and work days and 22.3 miles, numerosity literature would predict that the more persistently on a task; risk money and effort; and try to reach more numerous progress reported in miles should be perceived as round number benchmarks totally based on their arbitrary choice of higher than the less numerous progress measured in days. Yet, we scales. While number magnitude (numerosity) affects goal evalua- propose a potential reversal of this effect. We base our prediction on tions as shown in past literature, this effect is moderated and often another stream of research, which suggests that round numbers (end- reversed by the tendency to attain scale-induced round numbers. We ing with 0 or 5) serve as natural reference points to be achieved. It term this the scaleocity effect and further show that this tendency is has been shown that people try to reach round numbers as numerical mediated by the feeling of completeness stimulated by round num- goals in various domains, including SAT scores, baseball statistics, bers on specific measurement scales defining the progress. when trading currencies and stocks and running marathons (Allen, Overall, our research makes three important contributions to Dechow, Pope and George 2016; Pope and Simonsohn 2011; Osler the literature. First, we expand on the research on numerosity and 2003). While numerosity literature has often compared scales with unitosity (e.g., Burson, Larrick and Lynch 2009; Pandelaere, Bri- different granularity (e.g., months vs years) to demonstrate the role ers and Lembregts 2011; Monga and Bagchi 2012; Bagchi and Li of number magnitudes in consumer perceptions, the recent findings 2011; Cheema and Bagchi 2012) and integrate these efforts with the related to round numbers have been limited to use of single scales literature on round numbers and numerical cognition (e.g., Allen et and the tendency to reach the next round number. al. 2016; Pope and Simonsohn 2011; Lynn, Flynn and Helion 2013; The purpose of our research, therefore, was to examine scale- Bhattacharya, Holden, and Jacobsen 2012; Osler 2003; Gunasti and induced roundness and numerosity effects and show how use of Ozcan 2016; Gunasti and Devezer 2016; Yan and Duclos 2013) to different measurement scales affect perceptions and evaluations of introduce what we term the “scaleocity” effect. We demonstrate that progress and feeling of achievement. In a recent study Gunasti and people feel more accomplished when identical progress corresponds Ozcan (2016) found that round numbers create a sense of complete- to a round number on the specific scale they use even when com- ness, fullness or wholeness especially in brand names domain. Au- pared to a non-round number with a higher magnitude on another thors tested various real and fictitious brand names and product types scale. Second, we show that the scaleocity effect (scale-induced dif- to illustrate that consumers prefer brand names including numbers ferences in perceptions of progress) manifests itself as a tendency to ending with 0 and 5 over those with larger numbers because they per- spend more time and money, exert higher effort, persist longer in a ceive these products as more complete. This completeness effect was task, and take unnecessary risks just to reach a round number on an also investigated in a variety of different contexts such as complete- arbitrarily chosen scale (e.g., airlines miles). Third, we demonstrate

Advances in Consumer Research 799 Volume 45, ©2017 800 / The Role of Scales on Evaluations of Identical Goal Progress that this scaleocity effect is mediated by the perceived completeness Lynn, Michael, Sean Masaki Flynn, and Chelsea Helion (2013), induced by number roundness. “Do Consumers Prefer Round Prices? Evidence from Pay What-You-Want Decisions and Self-Pumped Gasoline REFERENCES Purchases,” Journal of Economic Psychology, 36 (June), 96- Allen, Eric J., Patricia M. Dechow, Devin G. Pope, and George Wu 102. (2016), “Reference-Dependent Preferences: Evidence from Monga, Ashwani, and Rajesh Bagchi (2012), “Years, Months, and Marathon Runners,” Management Science, Published Online: Days versus 1, 12, and 365: The Influence of Units versus April 20, 2016, [http://dx.doi.org/10.1287/mnsc.2015.2417] Numbers,” Journal of Consumer Research, 39 (1), 185-98. Bagchi, Rajesh, and Xingbo Li (2011), “Illusionary Progress Osler, Carol (2003), “Currency Orders and Exchange Rate in Loyalty Programs: Magnitudes, Reward Distances, and Dynamics: An Explanation for the Predictive Success of Step-Size Ambiguity,” Journal of Consumer Research, 37 Technical Analysis,” Journal of Finance, 58 (5), 1791–819. (February), 888-901. Ozcan, Timucin, and Daniel A. Sheinin (2012), “Effects of Bhattacharya, Utpal, Craig W. Holden, and Stacey Jacobsen (2012), Complete Products on Consumer Judgments,” Journal of “Penny Wise, Dollar Foolish: Buy–Sell Imbalances On and Product & Brand Management, 21 (4), 246-54. Around Round Numbers,” Management Science, 58 (2), 413- ______(2013), “Understanding Common Attribute Devaluation 31. in Multifunctional Products,” Journal of Marketing Theory Burson, Katherine, A., Richard, P. Larrick and John Jr. G. Lynch and Practice, 21 (4), 389-404. (2009), “Six of One, Half a Dozen of the Other: Expanding Pandelaere, Mario, Barbara Briers, and Christophe Lembregts and Contracting Numerical Dimensions Produces Preference (2011), “How to Make a 29% Increase Look Bigger: The Reversals,” Psychological Science, 20 (9), 1074-78. Unit Effect in Option Comparisons,”Journal of Consumer Cheema, Amar, and Rajesh Bagchi (2011), “The Effect of Goal Research, 38 (2), 308-22. Visualization on Goal Pursuit: Implications for Consumers and Pope, Devin, and Uri Simonsohn (2011), “Round Numbers as Managers,” Journal of Marketing, 75 (2), 109-23. Goals Evidence from Baseball, SAT Takers, and the Lab,” Gunasti, Kunter, and Timucin Ozcan (2016), “Consumer Reactions Psychological Science, 22 (1), 71-79. to Round Numbers in Brand Names,” Marketing Letters, 27 Yan, Dengfeng, and Rod Duclos (2013), “Making Sense of (2), 309-22. Numbers: Effects of Alphanumeric Brands on Consumer ______, and Berna Devezer (2016), ““How Competitor Brand Inference,” International Journal of Research in Marketing, Names Affect Within-Brand Choices,”Marketing Letters, 27 30 (2), 179-84. (4), 715-27. Meaning in Life, Powerlessness, and Nostalgia: Using Nostalgia to Mitigate the Effect of Powerlessness on the Search for Meaning in Life Jun Pang, Renmin University of China, China Sheng Bi, Washington State University, USA Huan Chen, Renmin University of China, China Andrew Perkins, Washington State University, USA

EXTENDED ABSTRACT talgic music and the other playing non-nostalgic music. A Chi-square Meaning in life refers to thoughts and feelings connected to test suggested a significant effect of power 2(χ (2) = 23.32, p<.001). the self-perceived significance of one’s own life (Steger and Frazier, Low-power participants were more likely to choose the nostalgic 2005). When one’s life meaning has been threatened, he/she will be concert than high-power participants (100% vs. 60%, χ2(1) = 24.12, motivated to restore their meaning in life. Previous research has es- p <.001) and the ones in the control condition (100% vs. 70.83%, tablished numerous threats that might undermine one’s perception of χ2(1)= 16.39, p <.001). No differences were found between the high- meaning in life, such as interpersonal rejection (Zadro, Williams, and power condition and the control condition (χ2(1) = 1.27, p =.18 ). Richardson 2004). The present research suggests an important and Study 3 aimed to test the underlying process. Forty eight par- very common threat: one’s feeling of powerlessness. ticipants were recruited. We first manipulated power by asking par- Individuals experience feelings of powerlessness in everyday ticipants to recall an experience in which they had high power or low settings. This ubiquitous feeling of powerlessness can result in an power. After that, participants indicated to what extent they wanted aversive experience associated with uncertainty, self-doubt, and neg- to search for meaning in life. Then they were asked to make a choice ative emotions (Galinsky, Gruenfeld, and Magee, 2003). To alleviate among four real snack products, two nostalgic and two non-nostalgic. this aversive experience, consumers tend to restore a sense of power A Chi-square test on the product choice showed a significant effect of by purchasing status-related products (Rucker and Galinsky, 2008). powerlessness on nostalgic preference (χ2(1)=3.46, p=.06). Next, we While the need to restore power is certainly an important individual used a bootstrapping approach to test the mediating process. Results motivator, we suggest that another motivator might be just as influen- indicated that after controlling for the effect of emotion (β = 0.25, p tial: restoring one’s meaning in life. Despite the ubiquity of instances = .32), low power led to a stronger motivation to search for meaning where one might feel powerless in society, as well as the power of the in life (β = -1.05, p = .002), which then increased the likelihood of search for meaning in life as a motivator, we are aware of no research choosing nostalgic products (β = .94, p = .03). Power had no direct that explore the relationship between feelings of powerlessness and effect on choice (β = -0.91, p =.23). Further, the mediating effect the search for meaning in life. The present research addresses this was negative and had a 95% confidence interval that excluded zero gap by examining whether feelings of powerlessness can influence (β = -.98, 95% CI=–3.29~–.10). These results support the proposed one’s motivation to search for and restore meaning in life. mediation process. One way an individual who feels powerless might attempt to Study 4 aimed to examine a boundary condition. We recruit- mitigate that feeling is through consumption activities. Prior research ed 176 participants from Mturk and randomly assigned them to 2 has shown that nostalgic products can activate consumers’ memories (power: low power vs. high power) × 2 (meaning in life: reading of the significant events they have experienced and enhance consum- vs. no reading) between-participants design. Participants were first ers’ perceptions of connectedness with important others (Wildschut assigned to either read an article that suggested that merely read- et al., 2010). We suggest that both of these experiences should help ing can provide meaning in life, or read nothing. Then participants consumers perceive their life as meaningful. Therefore, we predict were made to feel powerless or powerful by imagining that they are that the motivation to search for meaning in life resulting from feel- an employee or a boss. At last, they were asked to choose a soup ings of powerlessness will increase consumers’ preferences for nos- brand among the four real soup brands: two nostalgic brands and talgic products. two non-nostalgic brands. A logistic regression revealed a significant We conducted four studies to test these propositions. Study 1 power × reading interaction (χ2 = 5.76, p = .016). More importantly, measured the individual differences in power and examined its re- low-power participants who were not provided the article were more lationship with preference for nostalgic products. We recruited 111 likely to choose the nostalgic soup brands than high-power partici- participants at QQ Survey, a Chinese website similar Mturk. Par- pants (72.1% vs. 52.2%, χ2(1) = 3.74, p = .053). However, in the ticipants first completed a scale to measure their sense of power and reading condition, there was no significant difference. then reported overall attitudes toward an advertisement for instant In conclusion, the four studies provide consistent evidence that noodles that was manipulated to elicit feelings of nostalgia versus a consumers who feel powerless will prefer nostalgic products, and control. Results showed a significant power × nostalgia interaction (β that this effect is mediated by motivation to search for meaning in = .61, t(107) = 2.11, p = .037). As hypothesized, increased feelings life. These studies provide a new explanation for how individuals of powerless were positively related to preference for the nostalgic cope with feeling powerless. Beyond the theoretical contributions, noodle product (β = .44, 95% CI = .887, .002), but not the control (β this research also provides important implications on marketing = 0.17, 95% CI = –.199, .534). practice. Study 2 used a controlled experiment to test the main effect again. We recruited 196 participants at QQ Survey and randomly as- REFERENCES signed them to the high-power, the low-power, or the control condi- Galinsky, A. D., D. H. Gruenfeld, and J. C. Magee (2003), tion. We primed power firstly by asking participants to find out at “From Power to Action,” Journal of Personality and Social least 10 words from a 10 × 10 grid. The words were related to high Psychology, 85 (3), 453-466. power (e.g., boss and control), low power (e.g., subordinate and sub- Rucker, D. D. and A. D. Galinsky (2008), “Desire to Acquire: missive), or irrelevant to power (e.g., music and paper). Then partici- Powerlessness and Compensatory Consumption,” Journal of pants were asked to choose between two concerts, one playing nos- Consumer Research, 35 (2), 257-267.

Steger, Michael F and Patricia Frazier (2005), “Meaning in Life: Zadro, Lisa, Kipling D Williams, and Rick Richardson (2004), One Link in the Chain from Religiousness to Well-Being,” “How Low Can You Go? Ostracism by a Computer Is Journal of Counseling Psychology, 52 (4), 574-582. Sufficient to Lower Self-Reported Levels of Belonging, Wildschut, T., Sedikides, C., Routledge, C., Arndt, J., & Cordaro, F. Control, Self-Esteem, and Meaningful Existence,” Journal of (2010). Nostalgia as a repository of social connectedness: the Experimental Social Psychology, 40 (4), 560-567. role of attachment-related avoidance. Journal of personality and social psychology, 98(4), 573-586. The Asymmetric Effect of a Firm’s Invested Effort in the CSR Campaign Taehoon Park, University of South Carolina, USA Elise Chandon Ince, University of South Carolina, USA Anastasiya Pocheptsova Ghosh, University of Arizona, USA

EXTENDED ABSTRACT supporting multiple (vs. single) charities would be perceived as more An increasing number of companies are involved in prosocial effortful. Participants were randomly assigned to a 2 (mechanism behaviors called Corporate Social Responsibility (CSR). In contrast, type: insurance vs. response) by 2 (number of charities: multiple vs. firm-induced incidents that impair social good are coined Corporate single) between-subject design. Participants read about the same sce- Social Irresponsibility (CSI). This research investigates how con- nario as study 1 except for the effort manipulation. Then, participants sumers evaluate a company and its CSR campaign when the firm is reported how competent, effective, and efficient the company was involved in both a CSR and a CSI, depending on the order of the two: and the degree to which the company developed the campaign to the response (CSI-CSR) and the insurance (CSR-CSI) mechanism. offset the CSI incident. Finally, they rated the importance of having a Some work has demonstrated the positive evaluation of insurance CSR campaign for a company as a covariate. An ANCOVA revealed mechanism, because consumers attribute the cause of the negative a significant mechanism type by number of charity interaction on event to external factors if it is was preceded by CSR campaign perceived competence (F (1, 214) = 4.68, p = .032, = .021). In the (Klein and Dawar 2004). By contrast, another work (Wagner, Lutz, insurance condition, perceived competence was greater for the CSR and Weitz 2009) argues that the insurance mechanism is viewed campaign with the multiple (vs. single) activities, while perceived negatively when a CSI is highly congruent with a CSR, because competence did not differ in the response condition. A moderated consumers become suspicious about the firm’s true motive for CSR mediation supported our hypothesis: the indirect effect of the inter- campaign initiation. Expanding upon previous findings, we propose action between mechanism and number of charity through perceived a moderating factor that defines how consumers respond to the insur- motive was significant for the insurance condition (indirect effect = ance mechanism: invested effort by the firm in the CSR. -.11, SE = .08; 90% CI, -.28 to -.02), but not for the response condi- Consumers’ skepticism has been identified as an important faction (90% CI, -.06 to .02). tor that impacts how consumers evaluate a CSR campaign and the Study 3 investigates whether the source of CSR effort can firm (Drumwright 1996; Yoon, Gürhan-Canli, and Schwarz 2006). change firm’s evaluation. Specifically, when firms ask consumers to Although consumer suspicion about CSR motive can be derived get involved in choosing which charity to support (a common tactic from its temporal order in relation to a CSI incident (Wagner, Lutz, employed by companies), the effort comes from consumers and not and Weitz 2009), we argue that having a CSR campaign initiated the company, and therefore will not have a beneficial effect for the prior to a CSI does not harm company evaluation, when the firm insurance mechanism. The study employed 2 (mechanism type: in- invests high effort in the CSR campaign. Based on the previous find- surance vs. response) by 2 (charity choice: consumers vs. executives) ings that the level of firm’s effort in the campaign is interpreted as between-subject design. The scenario was about a hair care products the level of firm’s commitment to the social good (Ellen et al. 2000, company donates money to a non-profit organization supporting de- 2006), we argue that consumers will infer an altruistic motive from a velopment of low-carbon energy. For the source of effort, the com- CSR campaign with high effort in the insurance mechanism, improv- pany either let their consumers or executives choose the non-profit ing company evaluation. For the response mechanism, however, the organization. Participants reported their attitudes toward the compa- level of perceived ulterior motive will not differ based on the level of ny as well as the perceived motive of the campaign. An ANOVA re- invested effort due to that the motive of the campaign (i.e., to offset vealed a significant mechanism type by charity choice interaction on any wrongdoing) is clear. And thus, ironically, investing more effort firm evaluation (F (1, 158) = 6.49, p = .012, = .039). Participants in response to a CSI accident that has already happened will not im- in the insurance condition reported marginally lower evaluations of prove evaluations of the CSR campaign or the company. the firm with consumer (vs. executive) involvement. Interestingly, We present three studies testing our hypotheses. Study 1 inves- participants in the response condition reported marginally higher tigates whether more effort invested in a CSR campaign improves firm evaluation with consumer (vs. executive) involvement. Similar consumer reaction only in the insurance mechanism. The study em- pattern was found for perceived firm’s motive. A moderated media- ployed a 2 (mechanism type: insurance vs. response) by 2 (invested tion supported our hypothesis. effort: high vs. low) between-subject design. Participants read about Contributing to the knowledge of consumers’ response to com- a company that accidently polluted a local river. As the effort, the panies’ CSR efforts, this research suggests new ways for marketing company either donates money only or runs a task force team for riv- managers to mitigate the potential negative effects from engaging in er cleaning in addition to making a donation. Half of the participants CSR campaigns in light of CSI incidents. read that the CSR was initiated before the CSI, whereas the other half read the opposite order. Participants then reported the perceived REFERENCES product performance and expected campaign success. As expected, Drumwright, Minette (1996), “Company Advertising with a Social there was a significant mechanism type by invested effort interac- Dimension: The Role of Noneconomic Criteria,” The Journal tion on perceived product performance (F (1, 204) = 3.96, p = .048, of Marketing, 60(4), 71–87. = .019). Participants in the insurance condition reported margin- Ellen, Pam, Lois Mohr, and Deborah Webb (2000), “Charitable ally higher perceived product performance when the effort invested Programs and the Retailer: Do They Mix?,” Journal of by the firm was high (vs. low), while no difference emerged for the Retailing, 76(3), 393–406. response condition. Similar pattern was found for campaign success. Ellen, Pam, Deborah Webb, and Lois Mohr (2006), “Building The goal of study 2 was to replicate the effect with a different Corporate Associations: Consumer Attributions for Corporate manipulation of firm’s invested effort. As a proxy for invested effort, Socially Responsible Programs,” Journal of the Academy of the number of charities the firm supports was used expecting that Marketing Science, 34(2), 147–57.

Friestad, Marian and Peter Wright (1994), “The People Persuasion Yoon, Yeosun, Zeynep Gürhan-Canli, and Norbert Schwarz Cope with Knowledge Persuasion Model : How Attempts,” (2006), “The Effect of Corporate Social Responsibility (CSR) Journal of Consumer Research, 21(1), 1–31. Activities on Companies With Bad Reputations,” Journal of Klein, Jill and Niraj Dawar (2004), “Corporate Social Consumer Psychology, 16(4), 377–90. Responsibility and Consumers’ Attributions and Brand Evaluations in a Product-Harm Crisis,” International Journal of Research in Marketing, 21(3), 203–17. Wagner, Tillmann, Richard Lutz, and Barton Weitz (2009), “Corporate Hypocrisy: Overcoming the Threat of Inconsistent Corporate Social Responsibility Perceptions,” Journal of Marketing, 73(6), 77–91. Divided, Strategic, or Something Else? Exploring the Impact of Social Mobility on Habitus Erika L. Paulson, Quinnipiac University, USA

EXTENDED ABSTRACT intertwine with one another and attention is given to both constructs Bourdieu’s theory of distinction is commonly applied to social in the analysis. mobility. According to this theory, social mobility results in a divided Does the experience of social mobility conform more closely to habitus where individuals are forever caught between two compet- the idea of the divided habitus or the strategic habitus? Results sug- ing habituses. However, other have suggested that mobile individuals gest that the direction of mobility plays a crucial role in determining can pick and choose their habitus at will, a position termed a strate- the outcome. Furthermore, changes in the habitus do not always align gic habitus. Both positions have received empirical support, raising with either the divided or the strategic habitus. two questions. First, does the experience of social mobility conform For the downwardly mobile, habitus appears resistant to change. more closely to the idea of the divided habitus or strategic habitus? Adelaide’s story vividly illustrates downward mobility and a habitus Second, what factors influence the outcome? that remains upper class. She comes from an upper class family and Bourdieu’s theory of distinction is used to approach these re- attended exclusive private schools in her youth. However, the death search questions. Briefly, this theory works to understand how indi- of a parent meant a series of sudden, downward movements for her viduals draw boundaries between different social groups and compete throughout adolescence. In adulthood she married a working class to arrange them hierarchically. The currencies in this competition are man and the differences between Adelaide and her husband reveal social, economic, and cultural capital. Social class is believed to be that her habitus is still rooted in the upper class. She openly invokes mirrored in the amount of and type of capital possessed (Bernthal, her pedigree to explain why she has a different “level set” than her Crockett, and Rose 2005; Carfagna et al 2014; Coskuner-Balli and blue collar husband. She admits to “judging” his coworkers and Thompson 2012; Holt 1998; Moisio, Arnould, and Gentry 2013; bristles at working class children being fed Mountain Dew, sport- Üstüner and Holt 2007, 2010). ing buzzcuts, wearing camouflage, and using improper grammar. Her Cultural capital can be objectified in consumption items, dis- judgements indicate she retains her original habitus. This is because played through consumption behaviors and practices known as taste, there is little incentive for downwardly mobile individuals to adjust or embodied in the habitus. An individual’s habitus can be expressed to a less valued habitus. in a variety of ways including “standing, speaking, walking, […] feel- In contrast, the upwardly mobile exhibit more change to their ing and thinking” (Bourdieu 1990a, 70). Habitus is not only a char- habitus. It no longer matches their original habitus but neither does acteristic of an individual, but is also shared among those in a group it entirely conform to their new class. One example of such change such as a social class (Bourdieu 1990b, 77). What happens when an comes from Adam, the son of immigrant parents who have moved individual’s social class changes? Does their habitus change? steadily upward from the lower class. Overall, he describes himself Bourdieu theorizes that social mobility, either upward or down- falling somewhere in-between his parents and his new milieu. For ward, results in a divided habitus. Bourdieu describes how upwardly- example, Adam is “off-put” by his girlfriend’s ability to spend $300 mobile students “continually desire reintegration into their commu- or $400 in one shopping trip. Such feelings indicate his habitus has nity of origin but at the same time are also unable to fully assimilate not completely changed to fit his new class. For the upwardly mobile, into the elite” (Bourdieu 1998, 107). This outcome is known as a one’s habitus falls somewhere between their origin and destination. divided habitus or a habitus clivé and has received empirical support Upward and downward mobility have different effects on taste (Aarseth, Layton and Nielsen 2016; Lawler 1999; Saaticologu and because they are very different experiences. Downward mobility is Ozanne 2013; Savage 2015). embarrassing and provides no incentive for those who experience However, Lahire (2011) downplays the notion of a divided it to change their habitus. For the upwardly mobile the move up is habitus as one of many contradictions an individual experiences. celebrated, providing an incentive to change. It is also worth recon- Emmison (2003) further emphasizes how individuals make strate- sidering whether Bourdieu’s concept of a divided habitus accurately gic, tactical choices in which habitus to use. Such ideas have found captures the experience. Rather than reflecting double isolation, the empirical support (Abrahams and Ingram 2013; Demetry, Thurk, and habitus appears blended. Additional work must also consider the Fine 2015; Lehman 2009). Here, this position is referred to as a stra- role played by race, gender, and other characteristics as they inter- tegic habitus. sect with social mobility. Overall, this work provides a look into the A total of 26 life history interviews were conducted. Respon- experience of social mobility. Investigating the effects is particular- dents were first screened for subjective social mobility and semi- ly important given the frequent, dynamic nature of mobility in the structured questions were included measure mobility objectively. United States. The frequency of mobility also makes it essential to Respondents were nearly evenly composed of men and women. understand how mobility affects habitus and consumer behavior gen- Most were white, while others classified themselves as belonging erally. Exploring social mobility also highlights the need to revisit to one or more racial or ethnic groups. There was also a wide age traditional theories of the habitus. range and respondents came from every region of the United States. The grounded theory method was used to guide the analysis and in- REFERENCES terpretation of the interviews (Fischer and Otnes 2008; Glaser and Aarseth, Layton and Nielsen (2016), “Conflicts in the Habitus: The Strauss 1967; Goulding 2005; Spiggle 1994). Attention was given Emotional Work of Becoming Modern,” The Sociological to habitus as well as other closely related concepts. The habitus it- Review, 64 (1), 148-165. self has been subdivided into emotional (Arsel and Bean 2013) and Abrahams, Jessica, and Nicola Ingram (2013), “The Chameleon moral (Saatcioglu and Ozanne 2013) dimensions. While habitus and Habitus: Exploring Local Students’ Negotiations of Multiple consumption are conceptually distinct, in practice they overlap and Fields,” Sociological Research Online, 18 (4), 21.

Arsel, Zeynep, and Jonathan Bean (2013), “Taste Regimes and Goulding, Christina (2005), “Grounded Theory, Ethnography and Market-mediated Practice,” Journal of Consumer Research, 39 Phenomenology: A Comparative Analysis of Three Qualitative (February), 899-917. Strategies For Marketing Research,” European Journal of Bernthal, Matthew J., David Crockett, and Randall L. Rose (2005), Marketing, 39 (3/4), 294–308. “Credit Cards as Lifestyle Facilitators,” Journal of Consumer Holt, Douglas B. (1998), “Does Cultural Capital Structure Research, 32 (June), 130-145. American Consumption?” Journal of Consumer Research, 25 Bourdieu, Pierre (1984/2010), Distinction, trans. Richard Nice, (June), 1-25. London: Routledge. Lahire, Bernard (2011), The Plural Actor, trans. David Fernbach, Bourdieu, Pierre (1990a), The Logic of Practice, Cambridge: Polity Polity Press. Press. Lawler, Steph (1999), “’Getting Out and Getting Away’: Bourdieu, Pierre (1990b), In Other Words: Essays towards a Women’s Narratives of Class Mobility,” Feminist Review, 63 Reflexive Sociology, Cambridge: Polity Press. (September), 3-24. Bourdieu, Pierre (1998), The State Nobility: Elite Schools In The Lehmann, Wolfgang (2009), “Becoming Middle Class: How Field Of Power, Stanford: University Press. Working-Class University Students Draw and Transgress Carfagna, Lindsey B., Emilie A. Dubois, Connor Fitzmaurice, Moral Class Boundaries,” Sociology, 43 (August), 631-47. Monique Y. Ouimette, Juliet B. Schor, and Margaret Willis Moisio, Risto, Eric J. Arnould, and James W. Gentry (2013), (2014), “An Emerging Eco-Habitus: The Reconfiguration “Productive Consumption in the Class-Mediated Construction of High Cultural Capital Practices Among Ethical of Domestic Masculinity: Do-It-Yourself (DIY) Home Consumers,” Journal of Consumer Culture, 14 (July), 158-78. Improvement in Men’s Identity Work,” Journal of Consumer Coskuner-Balli, Gokcen, and Craig J. Thompson (2013), “The Research, 40 (August), 298-316. Status Costs of Subordinate Cultural Capital: At-Home Saatcioglu, Bige, and Julie L. Ozanne (2013), “Moral Habitus Fathers’ Collective Pursuit of Cultural Legitimacy Through and Status Negotiation in a Marginalized Working-Class Capitalizing Consumption Practices,” Journal of Consumer Neighborhood,” Journal of Consumer Research, 40 Research, 40 (June), 19-41. (December), 692-710. Demetry, Daphne, Jessica Thurk, and Gary Alan Fine (2015), Savage, Mike (2015), Social Class in the 21st Century, Pelican “Strategic Poverty: How Social and Cultural Capital Shapes Books. Low-income Life,” Journal of Consumer Culture, 15 (March), Spiggle, Susan (1994), “Analysis and Interpretation of Qualitative 86-109. Data in Consumer Research,” Journal of Consumer Research, Emmison, Michael (2003), “Social Class and Cultural Mobility: 21 (December), 491-503. Reconfiguring the Cultural Omnivore Thesis,” Journal of Üstüner, Tuba, and Douglas B. Holt (2007), “Dominated Consumer Sociology, 39 (September), 211-230. Acculturation: The Social Construction of Poor Migrant Fischer, Eileen, and Cele C. Otnes (2008), “Breaking New Ground: Women’s Consumer Identity Projects in a Turkish Squatter,” Developing Grounded Theories in Marketing and Consumer Journal of Consumer Research, 34 (June), 41-56. Behavior,” in Handbook Of Qualitative Research Methods In ----- (2010), “Toward a Theory of Status Consumption in Less Marketing, Russell W. Belk (Ed.), Elgar Publishing, 19–30. Industrialized Countries,” Journal of Consumer Research, 37 Glaser, Barry G. and Anselm L. Strauss (1967), The Discovery (June), 37-56. of Grounded Theory: Strategies for Qualitative Research, Chicago: Aldine. Self-Disclosure Asymmetry in Online Communities: A Challenge of Demographic Diversity Cornelia (Connie) Pechmann, University of California Irvine, USA Kelly EunJung Yoon, University of California Irvine, USA Denis Trapido, University of Washington Bothell, USA Judith J. Prochaska, Stanford University, USA

EXTENDED ABSTRACT daily reminders to post to their online community. Also, they were The success of online communities depends on the active en- encouraged to choose a quit date within one week of joining the on- gagement and strong attachment of the members (Ren et al., 2012). line community, so that all members were striving for the same goal Disclosure of personal information is a commonly proposed mecha- with similar milestones. nism that induces close interpersonal relationships (Ensari and Mill- The demographics of the online community members were er, 2002). Therefore, a question that arises is how online communities measured by a survey – gender, employment status, marital status, can encourage members to disclose themselves to the community, and age. Once members began to post to their online communities, and result in positive outcomes. Our study considers this question by the dyads that were formed were identified using social network examining online health communities for smoking cessation (Pech- analysis (details below). After this, the demographic similarity or mann et al., 2016). dissimilarity of each dyad was determined by comparing the dyad Self-disclosure is “any message about the self that a person members’ survey responses. Goal attainment, defined as sustained communicates to another” (Wheeless and Grotz, 1976, p. 47). Re- smoking abstinence, was assessed using email surveys (Pechmann searchers have found that disclosing personal information such as et al. 2016). A dyad was identified based on whether a member sent demographics is positively associated with developing high-quality at least one post to another member (Centola and van de Rijt 2015). relationships (Phillips et al., 2009). In addition, interpersonal simi- Tie strength was measured as the count of posts exchanged between larity in demographics helps individuals build close relationships two members of a dyad (Shriver et al. 2013). Individual-level self- (Naylor et al., 2012). However, consumers are often concerned about disclosure of demographics was assessed by two independent coders disclosing personal information to demographically dissimilar oth- who examined if a post discussed gender, employment status, marital ers, because this may potentially increase the psychological distance status, or age. Self-disclosure was coded in terms of occurrences and between them and o.thers (Phillips et al., 2009). This greater distance also content. Dyadic-level self-disclosure of examined demographics may undermine dyadic ties, which in turn may weaken online com- were coded as follows (Moon 2000): non-disclosure meaning neither munities dyad member self-disclosed the demographic (0), asymmetric self- Therefore, understanding how similarity in and self-disclosure disclosure meaning one member did so (1), and reciprocated self- of demographics may affect dyads seems important for helping on- disclosure meaning both members did so (2). line communities improve engagement and goal attainment. Hence, H1 predicted that demographically dissimilar versus similar we investigated how dyadic similarity in demographics affected dyads would show less reciprocated and more asymmetric self-dis- whether the members of the dyad disclosed their personal demo- closure of that demographic in their posts, if there was an obvious graphic information to others. We also examined whether dyadic minority on that demographic. To examine H1, our z-tests compared similarity affected the goal of dyadic abstinence (smoking cessation) the percentages of dissimilar versus similar dyads who reached each due to the mediating effects of dyadic disclosure, dyadic tie strength, stage of self-disclosure by demographic. The results supported the and total engagement. hypothesis. We expected to see dissimilarity-based self-disclosure asymmetry for gender and employment status because men and Hypothesis 1: In online communities, members of demographi- the unemployed had an obvious minority standing in these online cally dissimilar versus similar dyads will be (a) communities. Consistent with this expectation, gender dissimilar less likely to engage in reciprocated self-disclo- versus similar dyads were less likely to engage in reciprocated self- sure and (b) more likely to engage in asymmetric disclosure and more likely to engage in asymmetric self-disclosure self-disclosure about the demographic in their of gender in posts. Also, employment dissimilar versus similar dyads posts, but (c) only if there is an obvious minority were less likely to engage in reciprocated self-disclosure and more on that demographic. likely to engage in asymmetric self-disclosure of employment status in posts. For marital status and age, where there was no obvious mi- Hypothesis 2: In online communities, regardless of whether nority, these effects did not obtain. dyads are demographically similar or dissimilar, H2 posited that self-disclosure of either similar or dissimilar de- self-disclosure of demographics in posts will be mographics in posts would relate to positive dyadic outcomes. To test associated with tie strength. H2, self-disclosure type was regressed on each outcome, separately for similar and dissimilar dyads. For dissimilar dyads, most of the Our first study examined 8 online communities that participated results supported the hypothesis. For employment-dissimilar dyads, in a 100-day quit-smoking program called Tweet2Quit in 2012-2013 self-disclosure of employment status in posts related to tie strength. (Pechmann et al. 2016). Individuals who were interested in an online For martially dissimilar dyads, self-disclosure of marital status in community for quitting smoking and who met the screening criteria posts related to tie strength. For age dissimilar dyads, self-disclosure (i.e., ages 18-59, residing in the continental USA, current smokers, of age in posts related to tie strength. For similar dyads, self-dis- interested in quitting, mobile phone with data plan, social media us- closure of gender in posts related to tie strength. For employment- ers) were assigned to 20-member communities on Twitter consisting similar dyads, self-disclosure of employment status in posts related entirely of other smokers. Participants were mailed 8 weeks of study- to tie strength. For maritally similar dyads, self-disclosure of marital provided nicotine patches to help them quit and were sent automated status in posts related to tie strength.

In this research, we explored a psychological phenomenon, Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the self-disclosure asymmetry, wherein one member of a dyad self-dis- “like” button: The impact of mere virtual presence on brand closed his or her demographic but the other did not. We reasoned evaluations and purchase intentions in social media settings. that self-disclosure asymmetry might help to explain the weakness Journal of Marketing, 76(6), 105-120. of ties in online communities that are comprised of diverse strang- Pechmann, C., Delucchi, K., Lakon, C. M., & Prochaska, J. J. ers. Furthermore, we observed substantial self-disclosure inhibition (2016). Randomised controlled trial evaluation of Tweet2Quit: in the online communities that we studied, primarily in the form of a social network quit-smoking intervention. Tobacco control, asymmetric or one-sided self-disclosure. When there was an obvious tobaccocontrol-2015-052768. minority on a demographic, minority members of dyads chose to Phillips, K. W., Rothbard, N. P., & Dumas, T. L. (2009). To disclose conceal this demographic from majority members. or not to disclose? Status distance and self-disclosure in diverse environments. Academy of Management Review, 34(4), REFERENCES 710-732. Centola, D., & van de Rijt, A. (2015). Choosing your network: Ren, Y., Harper, F. M., Drenner, S., Terveen, L., Kiesler, S., Riedl, Social preferences in an online health community. Social J., & Kraut, R. E. (2012). Building member attachment in science & medicine, 125, 19-31. online communities: Applying theories of group identity and Ensari, N., & Miller, N. (2002). The out-group must not be so bad interpersonal bonds. Mis Quarterly, 36(3). after all: the effects of disclosure, typicality, and salience on Shriver, S. K., Nair, H. S., & Hofstetter, R. (2013). Social ties intergroup bias. Journal of personality and social psychology, and user-generated content: Evidence from an online social 83(2), 313. network. Management Science, 59(6), 1425-1443. Wheeless, L. R., & Grotz, J. (1976). Conceptualization and measurement of reported self‐disclosure. Human Communication Research, 2(4), 338-346. Advances in Consumer Research (Volume 45) / 809 The Conformity-Risk Paradox: Why Increasingly Risky Mortgages are Acquired by Increasingly Risk-Averse Consumers Léna Pellandini-Simányi, Università della Svizzera italiana, Switzerland Ádám Banai, Central Bank of Hungary, Hungary

EXTENDED ABSTRACT According to the ‘financialization of everyday life’ theory, con- We used a modified version of the question for investment risk sumers took increasingly risky mortgages before the crisis because tolerance of the American Survey of Consumer Finances: of the emergence of a new financial culture that promoted growing risk-tolerance (Aalbers 2008; Fligstein and Goldstein 2015). Rogers’ Imagine that you need a loan. Which of the following statements (2010) theory of the diffusion of innovation would imply a reverse on this card comes closest to the amount of financial risk that you are trend: that late adopters are more conformist and, by extension, more willing to take when you borrow money? risk-averse than early adopters. While in most markets product risk is constant or decreases over 1. Take substantial financial risk expecting to pay time (i.e. surgeries get safer as expertise accumulates, etc.), in some substantially lower installments markets, product risk increases. The mortgage market of the 2000s 2. Take above average financial risks expecting to pay below was one of such markets. In these markets, the conformity-based the average installments explanation implies a paradox relationship between risk-tolerance 3. Take average financial risks expecting to pay average and actual riskiness of products: increasingly risk-averse consumers installments should hold increasingly risky products simply because they jumped 4. Not willing to take any financial risk later on the bandwagon. We call this the conformity-risk paradox. We test this paradox through the case of mortgage borrowing, We further asked respondents how they would have answered examining two hypotheses: the above question at the time when they acquired their mortgage, and if it differed, used that answer in the analysis. Hypothesis 1: Increasingly risky mortgages were not taken by increasingly risk-tolerant consumers over time. 4. Borrower’s conformity

We define conformity as a tendency to follow other people’s Hypothesis 2: Increasingly risky mortgages were taken by in- behaviour and opinions as opposed one’s own judgement. People creasingly conformist consumers over time. may choose mortgages because their friends and acquaintances do so. We measured this aspect by asking respondents how many people Methodology they knew who held a similar mortgage at the time when they ac- Our test case is the Hungarian mortgage market between 2000 quired it (none= 0; one or two =1; many=2). Moreover, conformity and 2010, which meets the criteria of becoming increasingly risky is at play when people assess the riskiness of the mortgage based on over time (Balás et al. 2015). Mortgage borrowing was practical- the behavior and advice of others, such as friends, acquaintances or ly non-existent before 2000 in Hungary (Pellandini-Simányi et al. bank clerks, rather than their own assessment. We thus asked respon- 2015), hence the case also meets the criteria of being a new market. dents who considered their mortgage low-risk (more than 90% of the We used a survey with 189 mortgage borrowers, using face-to-face sample) to indicate the influences on their assessment (yes=1; no=0): interviews and random walk sampling method. The sample is representative of mortgage borrowers in Hungary in terms of age, educa- I considered the mortgage low-risk because… tion and settlement type (Balás et al. 2015).  many of my acquaintances, friends held a similar We used four variables: mortgage.  people whose opinion I respect suggested that it is not 1. Time of acquisition of the mortgage risky.

2. Actual mortgage risk The conformity score is the sum of these two, taking values between 0 and 2. We composed this variable based on three features: Analysis

(1) Denomination: based on the historical standard deviation Hypothesis 1: Increasingly risky mortgages were not taken by of exchange rates HUF = 1, EUR = 2, CHF =3 and other increasingly risk-tolerant consumers over time. FX (JPY loans) = 4. (2) Interest rate type: Variable =4, fixed=2. Our data shows a steady increase of riskiness of mortgages be- (3) Payment-to-Income ratio: below 10%=1, 10%-33%=2, tween 2000 and 2008 (from 5.5 to 8.4). Yet risk-tolerance remained 33%-50%=3, above 50%=4. consistently low: between 1.6 and 1.8 in all periods between 2003 and 2010, with even a minor, temporary decrease in 2005-6 (1= zero The riskiness score is the sum of the above, taking values from risk-tolerance, 2= average risk tolerance out of 4). Individual-level 3 to 12. differences between one’s self-assessed risk tolerance and the risk of one’s mortgage (labeled ‘risk attitude-behaviour gap’ (ABG)) shows 3. Borrower’s risk tolerance a steady upward trend from 4.2 to 6.4 between 2000 and 2008. These 810 / The Conformity-Risk Paradox: Why Increasingly Risky Mortgages are Acquired by Increasingly Risk-Averse Consumers findings confirm our hypothesis that increasingly risky borrowing increasingly risky products by increasingly risk-averse consumers, was not accompanied by an increase in risk tolerance. who jumped later the bandwagon.

Hypothesis 2: Increasingly risky mortgages were taken by in- REFERENCES creasingly conformist consumers over time. Aalbers, Manuel B. (2008), “The Financialization of Home and the Mortgage Market Crisis,” Competition & Change, 12 (2), Borrowers on average had a high number of acquaintances with 148–66. similar mortgages when they acquired their mortgage, suggesting an Balas, Tamás, Adam Banai, and Zsuzsanna Hosszú (2015), element of herd behavior. 90% of borrowers thought that their mort- “Modelling Probability of Default and Optimal Pti Level by gage was low-risk, mainly based on the behavior and opinion of oth- Using a Household Survey,” Acta Oeconomica, 65 (2), 183- ers. Conformity is high (above 0.9 between 2000 and 2004 and 1.2 209. between 2005 and 2008 on a scale of 0 to 2) throughout the period, Barsky, Robert B, F. Miles Kimball, Thomas Juster, and Matthew with a slight decrease in 2007-8. Analyzing conformity by risk cat- Shapiro (1997), “Preference Parameters and Behavioral egory shows that conformity tends to be higher for higher risk prod- Heterogeneity: An Experimental Approach in the Health and ucts, lending support to our hypothesis that increasingly conformist Retirement Study,” Quarterly Journal of Economics, 112 (2), consumers took the increasingly risky mortgages. This is supported 537-79. by the result that people of higher conformity levels have a signifi- Brown, Martin and Roman Graf (2012), “Financial Literacy, cantly higher risk attitude-behaviour gap (t-test statistics, significant Household Investment and Household Debt: Evidence from at p<0.05). To further test this finding, we run an ordered logit esti- Switzerland,” Working papers on finance No13/1,University of mation for the actual risk of the mortgage as dependent and the con- St.Gallen. formity index as independent variable, and gradually included risk- Brown, Sarah, Gaia Garino, and Karl Taylor (2013), “Household tolerance, age (at the time of borrowing), settlement type, education Debt and Attitudes toward Risk,” Review of Income and and income as control variables. Even after including all controls, a Wealth, 59 (2), 283-304. significant positive (0.529) relationship remained between - confor Fligstein, Neil and Adam Goldstein (2015), “The Emergence of a mity and the riskiness of the mortgage. Finance Culture in American Households, 1989-2007,” Socio- The widening attitude-behaviour gap has several alternative ex- Economic Review, 13 (3), 575-601. planations, which we excluded. First, it may be caused by a cohort Goldstein, Adam and Neil Fligstein (2014), “The Transformation effect: older people tend to be more risk-averse, hence an increase in of Mortgage Finance and the Industrial Roots of the Mortgage the average borrower age would also lead to decreasing risk-toler- Meltdow,” http://sociology.berkeley.edu/sites/default/files/ ance over time. T-test statistics, however, show that the average age faculty/fligstein/Rise%20and%20Fall%20Revised%20 (mid-thirties) of borrowers did not change significantly over the peri- June%202014%20Final%20Version.pdf. od. Second, the widening gap may be caused by the subprime exten- H. Peyton, Young (2015), “The Evolution of Social Norms,” sion of the market towards people of lower socio-economic status. Annual Review of Economics, 7, 359-87. Low SES status borrowers are more risk-averse (Pellandini-Simanyi Holló, Dániel (2009), “Risk Developments on the Retail Mortgage and Banai 2017); yet they only have access to higher risk financial Loan Market,” MNB Bulletin, 4 (3), 14-19. products. Thus, more low-SES borrowers entering the market over Hunter, Caroline and Judy Nixon (1999), “The Discourse of time may lead to a widening attitude-behaviour gap. This explana- Housing Debt: The Social Construction of Landlords, Lenders, tion, however, does not apply for the Hungarian case. The average Borrowers and Tenants,” Housing, Theory, and Society, 16 (4), income and labor market position of households taking mortgages 165-78. did not change significantly over the period; even, after 2007, high Langley, Paul (2008a), “Financialization and the Consumer Credit income households’ mortgage debt grew even quicker than that of Boom,” Competition & Change, 12 (2), 133-47. low income households (Tóth and Medgyesi, 2010; Balás, Banai and --- (2007), “Uncertain Subjects of Anglo-American Hosszú 2015). Financialization,” Cultural Critique, 65 (Fall), 67-91. --- (2008b), The Everyday Life of Global Finance: Saving and Conclusion Borrowing in Anglo-America, Oxford: Oxford University Theories of the proliferation of risky mortgages suggest a shift Press. in attitudes, either through the ‘financialization of everyday life’ Mandel, Naomi (2003), “Shifting Selves and Decision Making: The (Fligstein and Goldstein 2015) or through meaning-making process- Effects of Self-Construal Priming on Consumer Risk-Taking “ es that normalized risky mortgages (Peñaloza and Barnhart 2011). Journal of Consumer Research, 30 (1), 30-40. Our study shows that meanings need not change for consumers to Mendoza, Norma A. and John W. Pracejus (1997), “Buy Now, acquire risky products. Consumers did not see risky mortgages as in- Pay Later: Does a Future Temporal Orientation Affect Credit creasingly normal, but rather, mistakenly, as low-risk, inferred from Overuse?,” in Advances in Consumer Research, Volume 24, the behavior and opinion of others. Paradoxically, increasingly risky Vol. 24, ed. Merrie Brucks and Deborah J. MacInnis, Provo, mortgages were acquired by increasingly risk-averse consumers. UT: Association for Consumer Research, 499-503. Thus, instead of a shift in attitudes, the attitude-behavior gap opened Minsky, Hyman P. (1986), Stabilizing an Unstable Economy, New further over time. Haven and London: Yale University Press. Our findings are consistent with Roger’s (2010) diffusion of --- (1992), “The Financial Instability Hypothesis,” The Jerome Levy innovation theory, suggesting that less conformist and more risk- Economics Institute of Bard College, WP # 74. tolerant consumers adopt new products first, followed by more Pålsson, Anne-Marie (1996), “Household Saving Behaviour and conformist and risk-averse adopters. We extend this by arguing that Financial Management: Does the Degree of Relative Risk in (1) markets of increasingly risky offers and (2) high information Aversion Vary with Household Characteristics?,” Journal of asymmetry, this results in a conformity-risk paradox: the adoption of Economic Psychology, 17 (6), 771-87. Advances in Consumer Research (Volume 45) / 811

Pellandini-Simanyi, Lena and Banai, Adam (2017) “ Tversky, Amos and Daniel Kahneman (1985), “The Framing of Financialization of everyday life: Measurement and critique”, Decisions and the Psychology of Choice,” in Environmental Annual Meeting of the Society for the Advancement of Socio- Impact Assessment, Technology Assessment, and Risk Analysis, Economics, Lyon, France. ed. Mumpower J.L. Covello V.T., Stallen P.J.M., Uppuluri Pellandini-Simanyi Lena, Hammer Ferenc and Vargha Zsuzsanna. V.R.R., Berlin, Heidelberg: NATO ASI Series (Series G: (2015) “The Financialization of Everyday Life or the Ecological Sciences), vol 4. Springer. Domestication of Finance? How mortgages shape borrowers’ Vargha, Zsuzsanna (2011), “From Long-Term Savings to Instant temporal horizons, relationships and rationality in Hungary”. Mortgages: Financial Demonstrations and the Role of Cultural Studies, 59(5-6): 733-759. Interaction in Markets,” Organization, 18 (2), 215-35. Peñaloza, Lisa and Michelle Barnhart (2011), “Living U.S. Wainwright, Thomas (2009), “Laying the Foundations for a Capitalism: The Normalization of Credit/Debt,” Journal of Crisis: Mapping the Historico-Geographical Construction Consumer Research, 38 (4), 743-62. of Residential Mortgage Backed Securitization in the Uk,” Raghubir, Priya and Sanjiv R. Das (2010), “The Long and Short of International Journal of Urban and Regional Research, 33 (2), It: Why Are Stocks with Shorter Runs Preferred? ,” Journal of 372-88. Consumer Research, 36 (6), 964-82. Wennberg, John E., Benjamin A. Barnes, and Michael Zubkoff Rogers, Everett M. (2010), Diffusion of Innovations (4th Edition), (1982), “Professional Uncertainty and the Problem of New York: The Free Press. Supplier-Induced Demand,” Social Science & Medicine, 16 Schooley, Diane K. and Debra Drecnik Worden (1996), “Risk (7), 811-24. Aversion Measures: Comparing Attitudes and Asset Allocation,” Financial Services Review, 5 (2), 87-99. Tóth, István György and Márton Medgyesi (2010), “Jövedelemeloszlás és Megélhetési Nehézségek a Konszolidációs Csomagok és a Válságok Közepette Magyarországon,” in Közelkép. A válság munkapiaci hatásai, ed. György Molnár, Budapest: MTA Közgazdaságtudományi Intézet Országos Foglalkoztatási Közalapítvány, 180-200. Sadness Reduces Decisiveness Beatriz Pereira, Iowa State University, USA Scott Rick, University of Michigan, USA

EXTENDED ABSTRACT rant. We anticipated that sad participants would be more likely to se- People often make decisions when feeling distressed. Sadness lect the third option, which forces their friend to make the restaurant is a particularly prevalent and long-lasting emotion (e.g., Jordan et decision, even when doing so required sacrificing $10 in face value. al. 2011), and therefore understanding its impact on decision-making Results is particularly important. Recent research has found that sadness can We collapsed the first and second alternatives (French or Italian generate suboptimal decisions (e.g., increased willingness to pay and restaurant with a $60 face value), since both these options are equally impatience; Lerner et al. 2004, 2013). However, beyond quality of specific and choosing either require the same level of decisiveness. decision-making, sadness may influence the willingness to actively As predicted, Sadness participants were more likely to defer the engage in decision-making to begin with, an issue that has not been choice of restaurant to their friend (72%) than Neutral participants explored in research using forced-choice paradigms. (44%; χ2 (1) = 11.46, p = .001). Sadness increased indecisiveness According to appraisal theory (Smith & Ellsworth 1985), the when choosing for others, even when there is a cost ($10) associated way people cognitively appraise their environment is both a cause with it. and consequence of different emotions. Sadness, in particular, is characterized by a high sense of uncertainty (e.g., about one’s abil- Experiment 1B ity to cope; Tiedens and Linton 2001). Therefore, we propose that One of the limitations from Experiment 1A is that the gift-giv- sadness can impair decision-making by reducing decisiveness. In- ing context could be eliciting positive emotions, which could con- decisiveness may generate unnecessary search costs (Rassin et al. volute the interpretation of the results. Therefore, in experiment 1B 2008). Moreover, waiting too long to select popular options may re- we made the choice to be about a restaurant certificate to be used by sult in these becoming unavailable (students losing potential seats in the person making the decision. In addition, we expected sadness to a course, Palatano and Wengrovitz 2007). For consumers, indecisive- increase choice difficulty in this context, so we included such mea- ness can result in missed limited time offers opportunities. sure in this design. We hypothesized that the effect of sadness on Aside from costs to the consumer, indecisiveness can be costly decisiveness would be explained by the increased decision difficulty. for retailers. Purchase delays can increase stocking costs or turn into total revenue loss if the purchase is permanently deferred. Consum- Method ers’ difficulty in selecting a single alternative is one of the most im- One hundred sixty two participants were recruited from Ama- portant causes of delaying purchases (Greenleaf and Lehmann 1995). zon MTurk (53.1% female, mean age = 36.5). Participants watched Previous research has shown that negative affect integral to the the same movie clips from Experiment 1A. After the emotion manip- decision increases status quo choices (Luce 1998), but whether inci- ulation, all participants were told to imagine that they were going to dental sadness reduces decisiveness more generally is still an open purchase a restaurant voucher for an upcoming Restaurant Week in question. We demonstrate that the uncertainty that accompanies sad their hometown. The terms matched those of experiment 1A. Next, states increases perceived choice difficulty and, consequently, in- participants rated their choice difficulty (“How much difficulty did crease indecisive behavior. you experience when making the decision about which restaurant voucher to get?”). Experiment 1A Results One way that people can express decision avoidance is by cir- Participants in the Sadness condition were marginally more cumventing the responsibility for deciding (Anderson 2003). Hence, likely to select the more flexible, but lower value option (27%) than in Study 1A we examine whether incidental sadness increases inde- were participants in the Neutral condition (15%; χ2 (1) = 3.42, p = cision in the context of gift giving (e.g., whether gift-givers prefer .065). The choice shares of the flexible, but lower value option, were to buy a specific versus a generic gift card). In this context, there is smaller in the context of choice for the self, versus choice for oth- room for higher uncertainty given that people are inherently more ers (as expected, due to higher certainty regarding one’s own prefer- knowledgeable about their own preferences. ences). Method We ran regression analysis using PROCESS macro for SPSS One hundred forty one participants were recruited from Ama- (Hayes 2013) to test the hypothesis that choice difficulty mediates zon MTurk (44.7% female, mean age = 33.9). We induced either the effect of sadness on decisiveness. Sadness condition was a sig- sadness or a neutral emotional state with video clips validated in nificant predictor of choice difficulty (β = .614, SE = .22, p = .006). previous research (Gross and Levenson 1995, Lerner et al. 2004). Choice difficulty significantly predicted preference for the flexible Participants in the Sadness condition viewed a clip from The Champ option (β = .404, SE = .131, p = .002). Sadness is not a significant that portrays the death of a boy’s father. Participants in the Neutral predictor of flexible option when controlling for choice difficulty (β condition watched a video clip from a National Geographic docu- = .463, SE = .422, p = .273). The indirect effect of Sadness on choice mentary on coral reefs. of flexible option (i.e. indecisiveness) through choice difficulty was After the manipulation, all participants were told to imagine that significant and positive (β = .248, SE = .132, 95% CI [.0493, .6008], they were going to a birthday party, and that they had decided to give 1,000 bootstrap samples). Therefore, we conclude that sadness in- their friend a gift card. Their friend and their friend’s spouse love creases perceived choice difficulty, which in turn increases prefer- dining out, so the participant had decided to give them a restaurant ence for the flexible – and less decisive – option. gift certificate. Participants selected among Groupon gift cards (all costing $35): a $60 gift card to a French restaurant, a $60 gift card to an Italian restaurant, or a $50 gift card redeemable at either restau-

Advances in Consumer Research 812 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 813 Experiment 2 ness scores. The omnibus test was marginally significant (F(2, 201) Experiment 2 extends the findings of Experiment 1A and 1B = 2.50, p = .085). Pairwise comparisons show that indecisiveness in two ways. First, Experiment 2 sheds additional light on the hy- scores were significantly greater in the Sadness condition (M = 9.43, pothesized process (lack of certainty present when one experiences SD = 16.41) than in the Anger condition (M = 4.83, SD = 11.07; sadness) by manipulating anger, an emotion as aversive as sadness, t(134) = 1.99, p = .049) and the Neutral condition (M = 4.80, SD = but not high in uncertainty (Smith and Ellsworth 1985). Second, it 12.91; t(129) = 1.94, p = .054). Indecisiveness scores did not differ examines the influence of sadness on indecisiveness in a personal among Anger and Neutral conditions (p = .99). finance domain. Specifically, we gave participants a debt repayment Experiment 2 demonstrates that sadness reduces decisiveness, scenario where they could allocate all of their money to one of two while anger does not (relative to a Neutral control condition).These credit card accounts, or they could act more indecisively and split results suggest that not all negative emotions reduce decisiveness, the payment between accounts. We hypothesized that sadness would and that the appraisal of uncertainty that accompanies sadness is a increase the tendency to split one’s available money across cards potentially significant driver of the effect of sadness on decisiveness. (arguably, a proxy for indecisiveness), while anger would not since We replicated this finding in a follow-up correlational study (N = it is not associated with an uncertainty appraisal. 114, 55.3% female, mean age = 36.6) where participants rated their naturally occurring emotions and completed the debt repayment task Method (order counterbalanced). As predicted, sadness scores correlated We conducted a pre-test to test whether they perceived splitting positively and significantly with indecisiveness scores (r(112) = .33, the payment between accounts to be reflective of indecisive behav- p < .001). Indecisiveness did not correlate significantly with pride, ior. One hundred participants (recruited through MTurk) completed hope, indifference or happiness (all p’s > .10), but it correlated sig- the pre-test for a small payment (29% female, mean age = 32.5). nificantly with anger (r(112) = .20, p < .05). When both anger and They were asked to judge a target based on his/her decision in a sadness are entered in a multiple regression, only sadness predicts debt repayment scenario (Amar et al. 2011). Participants were asked indecisiveness (sadness: β = .211, t(100) = 2.86, p = .005; anger: to imagine that the target received $100 windfall from the govern- t(100) < 1). ment. The target had two debts where the money could be allocated: a Mastercard with a $100 balance and 10% APR and a Visa with a Experiment 3 $1,000 balance and 15% APR. Participants were randomly assigned Larger choice sets have been shown to increase decision avoid- to one of three conditions: the target (i) paid off the smallest ac- ance (Tversky and Shafir 1992), so if sadness reduces decisiveness, count, (ii) split payments evenly, or (iii) reduced the debt of the large, people experiencing sadness would be better off when facing smaller high-APR debt (the normative action). Participants rated the extent choice sets. Ironically, given that people experiencing sadness have to which they thought the target was feeling indecisive, uncertain, increased uncertainty over their preferences, they might believe that and risk-averse. We ran a repeated measures ANOVA to test whether they are better off by having larger assortments available. Therefore, participants perceived the three targets differently. Pairwise com- in Experiment 3 we tested whether sad participants would be willing parisons show that the target who split the funds between the cards to incur higher search costs to obtain a larger choice set. was perceived as being more indecisive, and uncertain. Importantly, this target was not perceived to be risk-averse, therefore splitting Method was not perceived as a strategy to reduce risk (which would conflict Two hundred seventy five students from a US Midwestern uni- with the proposed explanation that people split due to uncertainty). versity completed the experiment for course credit (39.6 female, Participants also stated that splitting was not a decision they would mean age = 21). We elicited sadness or a neutral state with the same have made, and did not believe that splitting made targets feel good. videos used in the previous experiments. Then, we gave participants Therefore, we conclude that participants perceived splitting as re- a short decision task where there is a trade-off between assortment flecting indecisiveness and uncertainty. and search costs. Participants are told that they are going to buy new Two hundred and four participants (MTurk) completed the main eyeglasses, and they found two stores online. From store A, one mile experiment (37% female, mean age = 32). We first induced sadness, away, they want to try 3 models. From store B, they want to try 15 anger, or neutral emotional states with video clips validated in pre- models, but store B is further. Participants are then asked how many vious research (Gross and Levenson 1995, Lerner et al. 2004, Rot- miles they are willing to drive in order to reach the store with the tenberg, Ray and Gross 2007). Participants in the Anger condition larger assortment (Inesi et al. 2011). viewed a four-minute clip from My Bodyguard that portrays a bully- Results ing incident. Participants were then presented with the scenario from Four participants with answers 3 standard deviations above the the pretest and asked to indicate how much of their $100 windfall mean were excluded from the analysis, two in each condition. As they would use to repay each card. The decision is difficult, because predicted, participants in the Sadness condition were willing to drive participants must choose between completely repaying the small bal- significant more miles to reach the store with larger assortment (M ance with a small APR (which is tempting) and chipping away at the = 21.34, SD = 15.49) than participants in the Neutral condition (M = high balance with a high APR (which is financially optimal). 17.60, SD = 12.86, t(269) = 2.15, p = .032). Although having a larger Results assortment will likely make their choice even more difficult (Tversky We computed an indecisiveness score, which captured the ex- and Shafir 1992), participants in the Sadness condition seem to be tent to which participants split their windfall evenly between the two coping with their uncertainty by having a stronger preference for the debts. Specifically, indecisiveness scores range from 0 to 50 and are larger assortment, even if this implies higher search costs. equal to min(Mastercard payment, Visa payment). If participants equally split their windfall between debts, their indecisiveness score General Discussion is min(50,50) = 50. If they allocate their entire windfall to one debt, Prior work has shown that sadness can either be detrimental their score is min(0,100) = 0. We ran a one-way between subjects or beneficial to decision-making, depending on the context and ANOVA to test whether emotional condition predicts indecisive- type of decision. Although this body of literature has documented 814 / Sadness Reduces Decisiveness directional effects of sadness on decision-making (e.g., increasing Gross, J. J., & Levenson, R. W. (1995). Emotion elicitation using impatience; Lerner et al. 2013), we propose and find across multiple films. Cognition and Emotion, 9(1), 87-108. experiments that sadness can actually reduce the propensity to be de- Hayes, A. F. (2013). Introduction to Mediation, Moderation, and cisive. We show that sadness increases indecisiveness in the context Conditional Process Analysis: A Regression-Based Approach, of financial decision-making regarding one’s current debts and in the New York, NY: Guilford Press. context of selecting gifts for others and for themselves. This effect is Inesi, M. E., Botti, S., Dubois, D., Rucker, D. D., & Galinsky, explained by higher levels of choice difficulty and uncertainty over A. D. (2011). Power and choice: Their dynamic interplay preferences. We also showed that sadness increases the propensity to in quenching the thirst for personal control. Psychological engage in more search costs to acquire a product, which exacerbates Science, 22(8), 1042-8. choice difficulty for people experiencing this emotion. Jordan, A. H., Monin, B., Dweck, C. S., Lovett, B. J., John, O. P., A potential limitation of the current findings is that the indeci- & Gross, J. J. (2011). Misery has more company than people sive behavior we observed could be interpreted as people seeking think: Underestimating the prevalence of others’ negative more variety (i.e., paying off more debt accounts, selecting the gift emotions. Personality and Social Psychology Bulletin, 37(1), card option that can be used in more restaurants). Arguably, the gift 120-35. card holder will still go to only one restaurant, and the debt holder Lerner, J. S., Li, Y., & Weber, E. U. (2013). The financial costs of will still have to pay both debts, so it is not clear that the flexible sadness. Psychological science, 24(1), 72-9. choices in our manipulations provide any variety in consumption. Lerner, J. S., Small, D. A., & Loewenstein, G. (2004). Heart strings Alternatively, choosing more variety might be a coping mecha- and purse strings: Carryover effects of emotions on economic nism of people experiencing indecisiveness, and not a desire to con- decisions. Psychological Science, 15(5), 337-41. sume more variety per se. Another possibility is that sad people may Luce, M. F. (1998). Choosing to avoid: Coping with negatively simply have an enhanced desire for multiple units (e.g., two small emotion-laden consumer decisions. Journal of consumer candies rather than one large candy). Multiple units may appeal to research, 24(4), 409-33. sad people for a number of reasons (e.g., being able to share with Patalano, A. L., & Wengrovitz, S. M. (2007). Indecisiveness others, or spreading out therapeutic consumption over time). Future and response to risk in deciding when to decide. Journal of research should also investigate this possibility. Behavioral Decision Making, 20, 405-24. Our work suggests that sadness reduces decisiveness due to in- Rassin, E., Muris, P., Booster, E., & Kolsloot, I. (2008). creased perceived choice difficulty, even when hesitation is costly. Indecisiveness and informational tunnel vision. Personality Alternatively, indecisiveness may prove beneficial when sadness- and Individual Differences, 45(1), 96-102. tinged decisions are suboptimal. It is possible that indecisiveness Rassin, E., Muris, P., Franken, I., Smit, M., & Wong, M. (2007). protects sad decision-makers from larger blunders. Measuring general indecisiveness. Journal of Psychopathology and Behavioral Assessment, 29(1), 60-7. REFERENCES Rottenberg, J., Ray, R. D., & Gross, J. J. (2007). Emotion elicitation Amar, M., Ariely, D., Ayal, S., Cryder, C., & Rick, S. (2011). using films. In J. A. Coan & J. J. B. Allen (Eds.), The Winning the battle but losing the war: The psychology of debt Handbook of Emotion Elicitation and Assessment (pp. 9-28). management. Journal of Marketing Research, 48, Special New York: Oxford University Press. Issue, S38-S50. Scherer, K. R., Wallbott, H. G., & Summerfield, A. B. (Ed.). Anderson, C. (2003). The psychology of doing nothing: (1986). Experiencing Emotion: A Cross-Cultural Study. Forms of decision avoidance result from reason and Cambridge University Press. emotion. Psychological Bulletin, 129, 139-67. Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive Bodenhausen, G. V., Gabriel, S., & Lineberger, M. (2000). appraisal in emotion. Journal of Personality and Social Sadness and susceptibility to judgmental bias: The case of Psychology, 48(4), 813-38. anchoring. Psychological Science, 11(4), 320-3. Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional Cryder, C. E., Lerner, J. S., Gross, J. J., & Dahl, R. E. (2008). certainty and uncertainty: The effects of specific emotions Misery is not miserly: Sad and self-focused individuals spend on information processing. Journal of personality and social more. Psychological Science, 19(6), 525-30. psychology, 81(6), 973-88. Dhar, R. (1997). Consumer preference for a no-choice option. Tversky, A., & Shafir, E. (1992). Choice under conflict: The Journal of Consumer Research, 24(2), 215-31. dynamics of deferred decision. Psychological Science, 3(6), Frost, R. O., & Shows, D. L. (1993). The nature and measurement 358-61. of compulsive indecisiveness. Behaviour Research and van Reekum, R., Stuss, D. T., & Ostrander, L. (2005). Apathy: Therapy, 31(7), 683-IN2. why care?. The Journal of Neuropsychiatry and Clinical Garg, N., & Lerner, J. S. (2013). Sadness and consumption. Journal Neurosciences, 17(1), 7-19. of Consumer Psychology, 23(1), 106-13. Verduyn, P., Delvaux, E., Van Coillie, H., Tuerlinckx, F., & Van Gentry, J. W., Kennedy, P. F., Paul, K., & Hill, R. P. (1994). The Mechelen, I. (2009). Predicting the duration of emotional vulnerability of those grieving the death of a loved one: experience: two experience sampling studies. Emotion, 9(1), Implications for public policy. Journal of Public Policy & 83-91. Marketing, 13(2) 128-42. Greenleaf, E. A., & Lehmann, D. R. (1995). Reasons for substantial delay in consumer decision making. Journal of Consumer Research, 22, 186-99. Market Memory Dolbec Pierre-Yann, Concordia University, Canada

EXTENDED ABSTRACT collective memories: what triggers and how memories are stored and Do markets “remember”? How do these memories shape mar- retrieved, the remaking of the past by the present by different actors, kets? Previous research has seldom examined the market processes the division between memorable and forgettable, that styles, genres, underlying the accumulation of memories around the evolution of a and discourses cannot be reduced to the aggregation of individual market. Instead, it has either concentrated on micro-level, consumer- subjectivities, the storification of the past and the associated simpli- centric processes (Biehal and Chakravarti 1982), or on how the inter- fication of memories as they are “prepared, planned, and rehearsed actions between other ‘types’ of memories, such as popular memories socially and individually”, and the shaping of reality by collective (Thompson and Tian 2008), nationalistic memories (Hartmann et al. memories (Olick 1999; Misztal 2003). 2016), and memories associated to specific brands (Brown, Kozinets, A third possible objection could be that a market memory is and Sherry 2003) can inform market processes such as myth making, similar to a commercial myth (e.g. Thompson and Tian 2008). Myths branding and market shaping. Moreover, a focus on the active work such as the South (Thompson and Tian 2008), the American Frontier of market actors in the evolution of markets (Ertimur and Coskuner- (Belk and Costa 1998), or the invisble hand of the market (Kennedy Balli 2015; Scaraboto and Fischer 2013) has left the role of market 2009) are the source of resonant meanings and ideals (Thompson devices, or “the material and discursive assemblages that intervene and Tian 2008) that “provide a logical model capable of overcoming in the construction of markets” (Muniesa, Millo and Callon 2007, 2) contradictions” (Levi-Strauss 1955, 443). They do so by providing a such as the memory of a market, out of the analysis. Studying such dramatized story that “selectively narrates … a version of the past,” devices on which the evolution of a market relies is crucial if we are flattening its complexity (Bell 2003, 76) and offering simplified, to further our understanding of market systems. As I will show, a packaged, and easily to understand narratives (Bell 2003). Through market memory is central in the evolution of a market by providing this process, memory events present in collective memories may be a shared yet conflictual representation of the past, structuring mar- amplified or silenced (e.g. Schwartz 2005). Myths and collective ket evolution through path dependencies, offering cultural resources, memories, albeit linked, are offering different representations of the shaping market actors’ identities, and linking a market with others. past. I broadly define a market memory as a collection of memory I analyzed an extensive archival dataset that retraces the evo- events that are collectively remembered by current market actors. By lution of fifteen established fashion houses (e.g., Balenciaga, Cha- memory event, I mean a certain event in time for which traces exist nel, Dior) representing four main subsets: ‘historical’ fashion house, (e.g., photographs, videos, press coverage). Such traces allow for the which have existed for more than 50 years, ‘established’ fashion activation, retrieval and re-interpretation of the memory event. My houses that have emerged in the last 25 years, ‘avant-garde’ designer findings are theoretically informed by two streams of research: work brands that have been recognized as revolutionizing the field, and in sociology on collective memories (e.g., Jedlowski 2001; Olick ‘new’ designer brands that have emerged in the last 10 years. For 1999) and work in management and institutional studies (e.g., Ante- these 15 fashion houses, I also collected the coverage of two years by and Molnar 2012; Lippman and Aldrich 2015; Ocasio, Mauskpf, (2010 and 2015) of fashion shows for the Fall and Summer seasons and Steele 2015). In both of these streams of research, memories from four main publications (New York Times, Vogue, Women’s are used to explain how a collectively shared and commemorated Wear Daily and Business of Fashion ). because fashion shows are past influences present events by “providing people with understand- intricately intertwined with the history of the luxury fashion market ings and symbolic frameworks that enable them to make sense of and because designers and critics alike reference previous memory the world” as we “rely on memory for the provision of symbolic events when creating or interpreting shows. Each memory event iden- representations and frames which can influence and organize both tified and deemed important in these shows was further researched to our actions and our conception of ourselves” (Misztal 2003, 13). This inquire about its origin, how it became salient (i.e., remembered by makes the study of a market memory central to our understanding of the market as an important memory event), and how it was stored and market systems. maintained over time. My dataset totals 819 pages. Let me prelude by discriminating a market memory from other I intertwine notions of memory work—“a conscious and pur- existing concepts. First, although the fields of history and memory poseful staging of memory” (Kuhn 2000, 186)—and institutional studies have collided and converged, a market memory is different work—actions that create, maintain, or disrupt the understanding, from the history of the same market. In line with existing work, I practices, and rules that govern a market (Dolbec and Fischer 2015; differentiate between history, which aims at providing if not an ob- Lawrence and Suddaby 2006)—to propose three types of memory jective at least a comprehensive if contentious account of the past work at the heart of the functioning of markets: (1) memory creation, (Poole 2008), and the retrieval, reinterpretation, and commemoration (2) memory maintenance, and (3) memory shaping, and explain how of the past which constitute collective memories (Schwartz 2005). these three types of memory work influences main institutional dy- More, history includes all that we know (or aim at knowing) about namics such as actors’ identities, institutional practices, other types a particular era, while collective memory emphasizes what market of institutional work, institutional logics, bases for legitimacy claims, actors remember (Poole 2008). Finally, collective memories also in- as well as interactions with other markets and how this defines a mar- clude “aspects of culture as well as social practices and structures” ket’s identity. (Reading 2003, 5). Memory creation influences the development of most institu- Second, the creation and collection of individual memories, tional dynamics by offering a bank of cultural resources for market although inextricably intertwined with that of collective memory actors to draw from and creating path dependencies. Perhaps one of (Olick and Robbins 1998) such as a market memory, are different. In- fashion’s most central memories is that of Chanel’s ‘little black dress’ dividual memories and their aggregation, which Olick (1999) refers (LBD). As the story goes, Vogue US published in 1926 a picture of to as collected memory, cannot address phenomena at the heart of a simple, long-sleeve, calf length straight black crepe de chine dress

Advances in Consumer Research 815 Volume 45, ©2017 816 / Market Memory by Chanel that they referred to as “Chanel’s Ford” as it could be- innovations, such as developing outfit that can take you from work come “a uniform for all women of taste” (Foreman 2014). If Chanel to the club, partnering with a major retail chain (JC Penny in 1983) introduced the little black dress, the garment got forever stamped in and artists (with Andy Warhol), his downfall was accompanied by consumers’ mind by its appearance on Audrey Hepburn in Breakfast his erasure from the fashion market memory. A number of reasons, at Tiffany. Even if this particular dress was designed by Givenchy, it from the downfall of his brand following the JC Penny collection to only furthered Chanel’s memory as the creator of the LBD and be- his ban as the creative director of his own brand to his death from a came a key memory event in the history of the designer, her fashion AIDS-related diseased in 1990s, could partly explain the desire of house, and fashion as a whole. As a cultural resource, the dress has the industry to distance itself from the world famous designer. been adopted and adapted by countless subcultures and designers, Finally, I explain how a market connects to adjacent markets by from mods to Dior’s ‘New Look’. And the dress now is an inescap- building linked market memories. This is important for the following able object, “a rite of passage for generations of designers” (Foreman two reasons: first, creating linkages with adjacent markets can rein- 2014). Its influence on the development of the fashion market has force or undermine market dynamics. For example, designers draw been such that it was included in the MoMA’s first fashion exhibi- from memory events from the markets of arts, architecture, music, tion that focuses on 99 items “that serve as a window onto social, and literature as referents for their collection. Not only does this al- economic and political changes in the world over the last 100 years.” low them to create symbolic associations for their work, but this in- Second, memory maintenance ensures the continuity of a mar- flux of memory events from adjacent markets that are structured by ket and associated dynamics by, for example, reifying the symbolic a similar institutional logic, the logic of art, also supports this logic power of memories and actors associated to them (e.g., Chanel’s in fashion. Second, linking memory events from adjacent markets little black dress constantly reaffirms the position of Chanel in fash- opens possibilities for market complexification. Work on institu- ion), and by stabilizing boundaries that keep market identities in tional theory has argued that markets are complex institutions where place. In fashion, the most efficient and lasting way to keep memo- firms are exposed to competing institutional demands (Greenwood et ries alive seem to be to have had a design piece consecrated at a al. 2011). I propose that these competing demands and contradicting staple of consumers’ everyday wardrobe (e.g. Chanel’s black dress, institutional resources and processes (Seo and Creed 2002) can favor Heidi Slimane’s slim suit). Memory maintenance is also performed institutional improvisation and market-level changes. through the use of objects, actors, events, places, practices, and firms My findings and the concept of market memory have a num- as points of reference. An example of this in the work of fashion crit- ber of theoretical implications. First, I offer a market device that ics is the coverage of the 2016 Fall-Winter menswear collection of can help explain the maintenance of markets and market changes Raf Simons by well-known critic Alexander Fury. In his article, Fury through memory contests. This points to the long-lasting value of references another fashion house (and associated designer), Maison any discursive act in markets, and how their accumulation, archival, Martin Margiela, mentioning that “the collection was so Margiela, in and retrieval can over time shape how a market develops. It posits distressing, its conspicuous wear, XXL-scale sweaters and coats sibly provides avenues for marketers to devise marketing strategies slipping and sliding off the figure”, that Raf Simons had “been fol- that aim at developing important memory events, or developing their lowing in Margiela’s tabi-toed footsteps all along—he’s previously memory-based skills to influence markets. Second, I further refine stated it was a Margiela show that triggered his interest in entering the link between institutional logics and collective memories (Oca- the industry.” Fury here uses Margiela as a cultural referent to ex- sio et al. 2016) by explaining how markets become intricately linked plain Simons’ collection, thereby reaffirming the place of Margiela, through references to memory events. Third, I also extend the con- the person and the fashion house, as well as key elements of its de- cept of collective memories to markets. This is an important step as sign, in the collective memory of the fashion market. previous work has positioned collective memories at the societal, Third, memory shaping allows for the reshaping of the past organizational, national, and communal levels. As markets as long- through forgetting and reinterpretation. This in turns leads to the lasting institutions that govern most of today’s social life, I redress re-configuration of the cultural resources available to market actors, this theoretical oversight. influencing for example market identities and sensemaking. Two types of memory shaping are of interest here. First, memory can be REFERENCES reinterpreted to reformulate how history unfolded. The reframing of Alexander, Ella (2011), “Halston,” Vogue, October 10th, http:// the origins of designers following their commercial success or the www.vogue.co.uk/article/roy-halston-frowick-biography. re-writing of the evolution of a fashion house following the outing Anteby, Michel and Virág Molnár (2012), “Collective Memory of its creative director are two examples of this process. For exam- Meets Organizational Identity: Remembering to Forget in a ple, Maison Martin Margiela, Yohji Yamamoto, and Rei Kawabuko Firm’s Rhetorical History,” Academy of Management Journal, are often presented in such a way that implies that their success and 55 (3), 525-540. recognition were immediate, although all designers suffered from Biehal, Gabriel, and Dipankar Chakravati (1982), “Information- poor sales and critical reception for their first few shows (e.g. Thur- Presentation Format and Learning Goals as Determinants man 2005), which helps reconcile their last influence on the fashion of Consumers’ Memory Retrieval and Choice Processes,” market. Second, memory erasure happens when memory events can Journal of Consumer Research, 8 (4), 431-441. undermine market dynamics. As other markets, the fashion market Brown, Stephen, Robert V. Kozinets, and John F. Sherry Jr. (2003), aims at shielding itself from some aspects of its functioning that “Teaching Old Brands New Tricks: Retro Branding and the could give rise to systematic critiques (e.g. Humphreys and Thomp- Revival of Brand Meaning,” Journal of Marketing, 67 (July), son 2014). If such critiques can arise from the ethical contradictions 19-33. at the heart of value chains, it can also emerge from negative memo- Dolbec, Pierre-Yann and Eileen Fischer (2015), “Re-Fashioning a ries that could undermine the identities of institutional actors. For Field: Connected Consumers and institutional Dynamics in example, even if Roy Halston was an iconic designer who shaped the Markets,” Journal of Consumer Research, 41 (6), 1447-1468. 1970s (Scheips 2013) whom Vogue dubbed “America’s first interna- tionally renowned designer” (Alexander 2011) and whom ideas and Advances in Consumer Research (Volume 45) / 817

Ertimur Burçak and Gokçen Coskuner-Balli (2015), “Navigating Olick, Jeffrey K. (1999), “Collective Memory: The Two Cultures,” the Institutional Logics of Markets: Implications for Strategic Sociological Theory, 17 (3), 333-348. Brand Management,” Journal of Marketing, 79 (2), 40-61. Olick, Jeffrey K. and Joyce Robbins (1998), “Social Memory Foreman, Katya (2014), “The Little Black Dress: Never Out of Studies: From Collective Memory to the Historical Sociology Style,” BBC, October 21st, http://www.bbc.com/culture/ of Mnemonic Practices,” Annual Review of Sociology, 24,105- story/20131209-the-dress-thats-always-in-style. 140. Fury, Alexander (2016), “Raf Simons,” Vogue, January 20th, http:// Phelps, Nicole (2016), “Vogue: Jean Paul Gaultier Spring 2016 www.vogue.com/fashion-shows/fall-2016-menswear/raf- Couture, ” Vogue, January 27th, http://www.vogue.com/ simons. fashion-shows/spring-2016-couture/jean-paul-gaultier. Hartmann, Benjamin J., Katja H. Brunk and Markus Giesler (2016), Reading, Anna (2003), “Editorial,” Media, Culture, and Society, 25, “Brand Retrofication,”Association for Consumer Research 5-6. 2016. Scaraboto, Daiane and Eileen Fischer (2013), “Frustrated Jedlowski, Paolo (2001), “Memory and Sociology: Themes and Fatshionistas: An Institutional Theory Perspective on Issues,” Memory and Society, 10 (1), 29-44. Consumer Quests for Greater Choice in Mainstream Markets,” Kuhn, Annette (2000), “A Journey through Memory,” in S. Journal of Consumer Research, 39 (6), 1234-257. Radstone (ed.) Memory and Methodology, New York, NY: Scheips, Charlie (2013), American Fashion, Assouline. Berg, 179-196. Schwartz, Barry (2005), “The New Gettysburg Address: Fusing Lawrence, Thomas B. and Roy Suddaby (2006), “Institutions and History and Memory,” Poetics, 33, 63-79. Institutional Work,” in S. R. Clegg, C. Hardy, T. B. Lawrence, Seo, Myeong-Gu and W. E. Douglas Creed (2002), “Institutional & W. R. Nord (Eds.) Handbook of Organization Studies, 2nd Contradictions, Praxis, and Institutional Change: A Dialectical Edition: 215-254. London: Sage. Perspective,” Academy of Management Review, 27 (2), 222- Lippman, Stephen and Howard Aldrich (2015), “A Rolling Stone 247. Gathers Momentum: Generational Units, Collective Memory, Thompson Craig J. and Kelly Tian (2008), “Reconstructing the and Entrepreneurship,” Academy of Management Review, South: How Commercial Myths Compete for Identity Value available at: http://amr.aom.org/content/early/2015/11/12/ through the Ideological Shaping of Popular Memories and amr.2014.0139. Countermemories” Journal of Consumer Research, 34 Misztal, Barbara A. (2003). Theories of Social Remembering. (February), 595-613. Philadelphia, PA: Open University Press. Thurman, Judith (2005), “The Misfit,”The New Yorker, July 4th, Moorman, Christine and Anne S. Miner (1998), “Organizational http://www.newyorker.com/magazine/2005/07/04/the-misfit-3. Improvisation and Organizational Memory,” Academy of Walsh, James P. and Gerardo Rivera Ungson (1991), Management Review, 23 (4), 698-723. “Organizational Memory,” Academy of Management Review, Muniesa, Fabian, Yuval Millo and Michel Callon (2007), “An 16 (1), 57-91. Introduction to Market Devices,” Sociological Review Monograph Series, Fabrian Muniesa, Yuval Millo and Michel Callon eds, 55 (2), 1-12. Ocasio, William, Michael Maskapf and Christopher Steele (2015), “History, Society, and Institutions: The Role of Collective Memory in the Emergence and Evolution of Societal Logics,” Academy of Management Review, available at: http://amr.aom. org/content/early/2015/11/12/amr.2014.0183.short. Curiosity and Want/Should Conflicts Evan Polman, University of Wisconsin-Madison, USA Rachel Ruttan, Northwestern University, USA Joann Peck, University of Wisconsin-Madison, USA

EXTENDED ABSTRACT empirical context that explores decision making and the want/should News and entertainment websites often increase online traffic conflict in consumers’ decisions, with an emphasis on how to recon- by tantalizing readers with sensational headlines containing phras- cile the conflict in a way that steers consumers toward more choice es such as: YOU WON’T BELIEVE WHAT HAPPENED NEXT, or of “should” options. YOU’LL BE SHOCKED WHEN YOU SEE THIS, or 17 SECRETS YOU DON’T WANT TO KNOW. Called clickbait, these headlines Want/Should Conflicts typically aim to exploit consumers’ curiosity, by providing just People regularly face a tension between what they want to do enough information to make readers curious, but not enough to sat- and what they believe they should do. Researchers have long stud- isfy their curiosity without clicking through to the linked content. ied the conflict between choosing options that provide immediate In a similar vein, many television episodes end on a cliffhanger – a gratification and options that are less desirable in the short-term, plot device featuring a precarious dilemma or shocking revelation at but provide more long-term benefits (e.g., Baumeister, Bratslavsky, the end of an episode – which spurs the audience to watch the next Muraven, and Tice 1998; Bazerman, Tenbrunsel, and Wade-Benzoni episode in order to find out how the story will continue. 1998; Haws 2016). Bazerman et al. (1998) dubbed this common This type of pursuit describes reconciling an information gap, struggle the “want/should conflict” whereby people face two com- or more precisely, what we call satisfying a curiosity appeal, which peting options. we define as a proviso that contains a promise to satisfy one’s piqued A common thread across past research is that making “should” curiosity provided one option is chosen over others. In the current choices is an effortful decision, requiring individuals to both override research we leverage the motivational benefits of curiosity appeals immediately rewarding temptations and to enact willpower (Milk- to test interventions designed to help steer people away from tempt- man, Rogers, and Bazerman 2008). Accordingly, much of the litera- ing “want” options like choosing unhealthy foods, choosing smaller- ture to date has focused on how to increase consumers’ willpower. sooner payouts, watching lowbrow films, taking the elevator, and For example, increasing individuals’ self-control capacities and cog- toward less-than-tempting, though normatively desirable “should” nitive resources are well-studied psychological tools that enhance the options. In all, our interventions leverage curiosity by luring people ability to make “should” choices (e.g., Baumeister et al. 1998; Shiv with the information that satisfies a curiosity appeal. To illustrate, and Fedorikhin 1999). However, another approach is to decrease the imagine having the solution to a puzzling trivia question revealed to level of temptation evoked by “want” options (Hoch and Lowenstein you if you choose to take the stairs rather than take the elevator. By 1991). For instance, providing individuals with some sense of imme- promising satisfaction to the curiosity appeal (caused by the trivia diate gratification (that is separate from what the “want” option itself question) with choosing the “should” option (taking the stairs), we provides) can reduce the immediate desire for a “want” option in propose the “should” option is more likely to be chosen than when favor of a “should” option. In one empirical demonstration, Urmin- there is no curiosity appeal (and corresponding motivation to satisfy sky and Kivetz (2011) provided people with an immediate financial it). bonus in an intertemporal choice between a smaller-sooner option Our approach to motivating consumers via curiosity appeals and a larger-later option. The authors found that the bonus led people implies that curiosity appeals could be used to motivate consum- to prefer the larger-later option because receiving it provided a sense ers to choose all kinds of options, besides solely “should” options. of immediate gratification, thus making it easier to choose the less However, we also show that when curiosity appeals are misapplied tempting but higher valued “should” option. In the current research, to “want” options, consumers will counter-intuitively choose “want” we explore whether appealing to a positive, intrinsically motivating options with less (not more) likelihood. At first glance, it might seem state (curiosity) that requires no financial incentives could similarly like nudges (and curiosity appeals in particular) would hypothetical- steer people toward more “should” options. ly operate in the same way across options – for example, in a choice between option A and B, wherein option A is nudged, it should make Curiosity little difference if option A is a “should” option or a “want” option Research has found that curiosity is a powerful predictor of be- (at a minimum, the nudge ought to increase options A’s likelihood in havior, playing an important role in motivating learning, mastery, and choice regardless of whether it is a “should” or “want” option). How- facilitating scientific discovery (Koestler 1964). Curiosity involves ever, we show that nudging a “want” option via a curiosity appeal is positive feelings of interest, but also feelings of uncertainty due to a a heretofore undocumented and novel way to increasing consumers’ perceived lack of knowledge (Litman and Jimerson 2004). An impor- choice of “should” options. That is, in our research we evidence two tant theory of curiosity suggests that curiosity signals the presence of different ways for using curiosity to change consumers’ decisions an “information gap” – that is, a lack of desired experience or knowl- and increase their choice of “should” options: one where “should” edge (Loewenstein 1994; see also Litman 2005; Menon and Soman options are nudged, and another where “want” options are nudged. 2002). This feeling of deprivation instills a motivation to seek out Our research is important because a large portion of marketing the missing information in order to reduce or eliminate the feeling of is about grabbing consumers’ attention, and curiosity is a feeling that deprivation (Maner and Gerend 2007), even if the missing informa- activates attention (Gottlieb, Oudeyer, Lopes, and Baranes 2013). tion is unpleasant (Kruger and Evans 2009) or causes people physical Yet, empirical research on curiosity and its relevance to marketing pain when they try to resolve their curiosity (Hsee and Ruan 2016). has been slow to keep up, with a few notable exceptions such as Isik- Despite this feeling, curiosity is not typically seen as an aversive man, MacInnis, Ülkümen, and Cavanaugh (2016) and Menon and state; on the contrary, curiosity has been found to mitigate a negative Soman (2002). In the current research, we examine curiosity in an experience, by improving upon it (Isikman et al. 2016); and Loew-

Advances in Consumer Research 818 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 819 enstein (1994) remarks that people like to make themselves curious indulgence, by eschewing too much indulgence (Dhar and Simonson precisely to satisfy their curiosity. Indeed, positive feelings such as 1999). Thus, while choosing a “should” and being rewarded with the novelty, surprise, and closure help define curiosity: a desire to know satisfaction of one’s curiosity feels like a balanced choice, choosing something (Gottlieb et al 2013). a “want” and being rewarded with the satisfaction of one’s curiosity Drawing on research supporting the motivational power of cu- feels like an overindulgence, and is consequently avoided. riosity, we tested a previously unexplored strategy for steering behavior by creating interventions that discourage tempting behaviors Studies (“wants”) and instead encourage less-than-tempting but nonetheless In study 1, we gave participants a choice between two fortune normatively desirable behaviors (“shoulds”). Our central prediction cookies: one plain and one dipped in chocolate and covered in sprin- was that, in a choice between “want” and “should” options, a cu- kles. Half the participants were given no additional information – riosity appeal that can be satisfied by selecting a “should” option and this control group chose the cookies with a 20-80 split in favor of will increase the choice of “should” options over “want” options. the chocolate-dipped cookie. The other half comprised our treatment From a theoretical standpoint, it is noteworthy that no research has condition. We told participants in this group that the plain cookie yet investigated how curiosity and temptation jointly influence con- contained a fortune that would tell them something personal that we sumer behavior; nor has research investigated whether receiving knew about them. This undoubtedly piqued their curiosity, and we satisfaction to one’s curiosity will change, much less reverse one’s observed a near complete reversal of preferences. choices or preferences. To be sure, research has found that curiosity- In study 2, we examined the special case that describes nudg- inducing advertising has the effect of making consumers search for ing a “want” option with a curiosity appeal. Thus, study 2 has three information (Krugman 1965); and that consumers tend to remember conditions – the same two conditions from study 1, plus a condition a curiosity-inducing brand better (Menon and Soman 2002) – but where a “want” option is nudged via a curiosity appeal. We found no research has looked at the effect of curiosity on making an ac- that in an intertemporal choice, over 80% of participants in the con- tual choice, precisely one of the major aims of consumer behavior trol condition chose the smaller-sooner “want” option. Yet when we research. tied participants’ choices to piquing their curiosity (with revealing the answers to a self-assessment quiz that participants took), the pro- Nudging “Wants” portion of participants choosing the “want” option dropped signifi- Implied by our research, curiosity appeals can be used to nudge cantly, in both instances. These two separate paths to increasing par- consumers to choose other options, besides solely “should” options. ticipants’ choice of “should” options provide evidence that curiosity For instance, curiosity appeals (and nudges writ large) could con- appeals do more than simply incentivize participants into choosing ceivably be used to nudge consumers to choose “want” options. De- one option over another, because we found that when “wants” are spite the voluminous research on nudges, the notion that they can be nudged with curiosity appeals, participants are disincentivized by the misused to increase “want” options among consumers has not been promise of curiosity relief. investigated. However, we propose that when “wants” are nudged In study 3, we investigated the different processes that underlie (via curiosity appeals), consumers will counter-intuitively prefer why people choose a “should” when it is nudged, and why people them less. also choose a “should” when a “want” is nudged. To do this, we Prior research has shown that before consumers choose “want” measured participants’ dispositional curiosity, and we measured par- options, they prefer to feel like they have earned the right to choose ticipants’ unease, guilt, and overindulgence with choosing a nudged them (Kivetz and Simonson 2002). This is because consumers often “want” option. We found that participants were more likely to choose feel guilty and uneasy when choosing “want” options (Kivetz and Si- a highbrow (“should”) film clip over a lowbrow (“want”) film clip monson 2002; Mishra and Mishra 2011; Ramanathan and Williams when we piqued their curiosity (with a magic trick) and promised to 2007). As such, consumers feel like they require a reason (some jus- satisfy their curiosity (revealing the secret behind the trick) provided tification) for choosing “want” options (Okada 2005). In support of they choose the highbrow film. This tendency was stronger among this view, research has shown that bolstering one’s self-concept or participants with higher dispositional curiosity (while holding con- bolstering one’s effort serve as guilt-reducing ways that help make stant other feelings such as fun, novelty, and attention). In a separate “want” options more justifiable (Khan and Dhar 2007; Kivetz and bootstrapping analysis, we found that when the “want” film clip is Simonson 2002). In fact, quite often with compensatory decision nudged, it makes people feel guilty choosing it, hence they prefer the making, the path to justifying and choosing “want” options tends alternative “should” film clip. For results to both tests (importantly, to involve first choosing “should” options – as though consuming we did not find evidence supporting the possibility that psychologi- “should” options provides consumers with the feeling that they have cal reactance is driving the effects). earned the license to subsequently indulge and choose “want” op- Then, in studies 4 and 5 we broadened the scope and ecological tions (Merritt, Effron, and Monin 2010). validity of our investigation by moving to the field. In study 4, in a Thus, when consumers have established that they have earned building on a large university campus, we designed a placard with an the right to indulge, they feel like they have leeway for choosing unanswered trivia question that we placed by the elevators. Then in a a “want” option. Nudges, however, are designed precisely to help nearby stairwell, we placed different placards with the answers to the consumers make decisions without having to invoke extensive ef- question (thus satisfying their curiosity). We changed the questions fort (Thaler and Sunstein 2009). And without the feeling of spending and corresponding answers every day, and ran our experiment for effort, consumers may not feel like they have adequately earned the 28 days. The first 14 days made up our pretest phase. In this phase, right to indulge. What is more, satisfying one’s curiosity evokes an we measured foot traffic (with a SenSource people-counting laser) overall positive feeling – it serves as a reward in itself (Loewen- in the intervention stairwell without our placards. In addition, we stein, 1994). When this feeling of reward is coupled with choosing measured foot traffic in a separate stairwell on the other side of the a “want” option, people may feel like they are, in the moment, over- building (our so-called control stairwell). The test revealed a signifi- indulging if they choose a “want” option. As the literatures on bal- cant interaction, such that the number of counts in the intervention ancing and hedonic editing show, consumers prefer to balance their stairwell during the posttest significantly exceeded the number of 820 / Curiosity and Want/Should Conflicts counts during the pretest phase, whereas the number of counts in Khan, Uzma, and Ravi Dhar (2007), “Where there is a Way, the control stairwell during the posttest and pretest phases was not Is There a Will? The Effect of Future Choices on Self- significantly different. Control.” Journal of Experimental Psychology: General, 136 In Study 5 we partnered with a locally-owned grocery store and (2), 277-288. tested whether curiosity gaps can motivate people to buy more fruits Kivetz, Ran, and Itamar Simonson (2002), “Earning the Right to and vegetables. For each of the produce items (e.g., beets) we cre- Indulge: Effort as a Determinant of Customer Preferences ated a placard with a joke on it (e.g., “why did the beets blush?”) Toward Frequency Program Rewards,” Journal of Marketing and posted the placard by the produce item’s regular sign that con- Research, 39 (2), 155-170. tained its details like price and origin. Then, close to the placard, we Koestler, Arthur (1964), The Act of Creation. London: Hutchinson. placed a cup containing bag closures with the respective punchline Kruger, Justin and Matt Evans (2009), “The Paradox of Alypius printed on them (e.g., “because the beets saw the salad dressing”). and the Pursuit of Unwanted Information,” Journal of The placard indicated the punchline is on the nearby bag closures. In Experimental Social Psychology, 45 (6), 1173-9. all, we had 17081 cases, and we found a significant 10% increase in Krugman, Herbert E. (1965), “The Impact of Television fruits and vegetables purchases when our curiosity intervention was Advertising: Learning Without Involvement,” Public Opinion imposed. Quarterly, 29 (3), 349-356. In sum, our research demonstrates a new phenomenon in want/ Litman, Jordan (2005), “Curiosity and the Pleasures of Learning: should conflict that goes beyond making “should” options easier to Wanting and Liking New Information,” Cognition & Emotion, choose. Although we found that nudging “shoulds” increases the 19(6), 793-814. extent that they are chosen, we also found that nudging “wants” in- Litman, Jordan A. (2008), “Interest and Deprivation Factors creases the extent that “should” options (rather than “want” options) of Epistemic Curiosity.” Personality and Individual are chosen. Taken together, our research shows that the powerful ef- Differences, 44 (7), 1585-1595. fects of curiosity come at a small cost and help promote a wide range Litman, Jordan A. and Tiffany L. Jimerson (2004), “The of desirable behaviors, from choosing high-brow films to exercising Measurement of Curiosity as a Feeling of Deprivation,” more to healthier eating. Journal of Personality Assessment, 82 (2), 147-57. Loewenstein, George (1994), “The Psychology of Curiosity: A REFERENCES Review and Reinterpretation.” Psychological Bulletin, 116(1), Bastardi, Anthony and Eldar Shafir (1998), “On the Pursuit and 75. Misuse of Useless Information.” Journal of Personality and ——— (1996), “Out of Control: Visceral Influences on Behavior,” Social Psychology, 75 (1), 19-32. Organizational Behavior and Human Decision Processes, 65 Baumeister, Roy F., Ellen Bratslavsky, Mark Muraven, and Dianne (3), 272-92. M. Tice (1998), “Ego Depletion: Is the Active Self a Limited Maner, Jon K. and Mary A. Gerend (2007), “Motivationally Resource?” Journal of Personality and Social Psychology, 74 Selective Risk Judgments: Do Fear and Curiosity Boost the (5), 1252-1265. Boons or the Banes?” Organizational Behavior and Human Bazerman, Max H., Ann E. Tenbrunsel, and Kimberly Wade- Decision Processes, 103 (2), 256-67. Benzoni (1998), “Negotiating with Yourself and Losing: Merritt, Anna C., Daniel A. Effron, and Benoît Monin (2010), Making Decisions with Competing Internal Preferences,” “Moral Self‐Licensing: When Being Good Frees us to be Academy of Management Review, 23 (2), 225-41. Bad,” Social and Personality Psychology Compass, 4 (5), Berlyne, Daniel E. (1960), Conflict, Arousal, and Curiosity. 344-357. McGraw-Hill Book Company. Menon, Satya and Dilip Soman (2002), “Managing the Power of Bitterly, T. Bradford, Robert Mislavsky, Hengchen Dai, and Curiosity for Effective Web Advertising Strategies,”Journal of Katherine L. Milkman (2015), “Dueling with Desire: A Advertising, 31 (3), 1-14. Synthesis of Past Research on Want/Should Conflict,” inThe Milkman, Katherine L., Todd Rogers, and Max H. Bazerman Psychology of Desire, W. Hofmann and L. Nordgren, eds. (2008), “Harnessing our Inner Angels and Demons: What Dhar, Ravi, and Itamar Simonson (1999), “Making Complementary we have Learned about want/should Conflicts and how that Choices in Consumption Episodes: Highlighting Versus Knowledge can Help Us Reduce Short-Sighted Decision Balancing,” Journal of Marketing Research, 36 (1), 29-44. Making,” Perspectives on Psychological Science, 3 (4), 324- Gottlieb, James, Oudeyer, Pierre-Yves, Lopes, Manuel, and Adrien 38. Barnes (2013), “Information-seeking, curiosity, and attention: Mishra, Arul, and Himanshu Mishra (2011), “The Influence of Price Computational and neural mechanisms,” Trends in Cognitive Discount versus Bonus Pack on the Preference for Virtue and Sciences, 17 (11), 585-593. Vice Foods,” Journal of Marketing Research, 48 (1), 196-206. Haws, Kelly L. (2016), “Enhancing Self-Control in Consumer Okada, Erica Mina (2005), “Justification Effects on Consumer Decisions,” Current Opinion in Psychology, 10, 118-123. Choice of Hedonic and Utilitarian Goods,” Journal of Hoch, Stephen J., and George F. Loewenstein (1991), “Time- Marketing Research, 42 (1), 43-53. Inconsistent Preferences and Consumer Self-Control.” Journal Ramanathan, Suresh, and Patti Williams (2007), “Immediate of Consumer Research, 17 (4), 492-507. and Delayed Emotional Consequences of Indulgence: Hsee, Christopher K. and Bowen Ruan (2016), “The Pandora The Moderating Influence of Personality Type on Mixed Effect: The Power and Peril of Curiosity,” Psychological Emotions.” Journal of Consumer Research, 34 (2), 212-223. Science. Shiv, Baba, and Alexander Fedorikhin (1999), “Heart and Mind in Hsee, Christopher K., Yang, and Bowen Ruan (2015), “The Mere Conflict: The Interplay of Affect and Cognition in Consumer Reaction Effect: Even Non-Positive and Non-Informative Decision-Making,” Journal of Consumer Research, 26 (3), Reactions can Reinforce Actions,” Journal of Consumer 278-292. Research. Advances in Consumer Research (Volume 45) / 821

Thaler, Richard H. and Cass R. Sunstein (2009), Nudge: Improving Urminsky, Oleg, and Ran Kivetz (2011), “Scope Insensitivity and Decisions about Health, Wealth, and Happiness. Penguin the “Mere Token” Effect.” Journal of Marketing Research, 48 Books. (2), 282-295. Thaler, Richard H., and Eric J. Johnson (1990), “Gambling with the House Money and Trying to Break Even: The Effects of Prior Outcomes on Risky Choice,” Management Science, 36 (6), 643-660. Evaluation Overconfidence: When Uncertainty in Attribute Under Produces Less Extreme Product Evaluations Deidre Popovich, Texas Tech University, USA Ryan Hamilton, Emory University, USA

EXTENDED ABSTRACT knowledge, such that their confidence was well founded, the knowl- Consumers’ ability to translate common product attributes edge probe did not lead to less extreme evaluations (Experiment 4). (watts, calories, megapixels) into subjective evaluations (perceived This phenomenon is robust across various manipulations of energy efficiency, perceived healthiness, perceived picture quality) knowledge probes, including explaining an attribute (Pilot Experi- is fundamental to theories of consumer choice. One factor that might ment), making relative estimation (Experiment 1), estimating the affect these evaluations is consumers’ confidence in their- under focal option’s attribute level (Experiments 2-4), and estimating the standing of these commonly encountered attributes. This research attribute level of unrelated options (Experiment 5). Taken together, proposes that consumers are often overconfident in assessing their the effect is consistent: a knowledge probe, regardless of the type, own understanding. We show that simple interventions that encour- caused less extreme product evaluations. age consumers to reflect on the level of their knowledge (e.g., es- This research provides several important contributions to theo- timating an attribute level, explaining an attribute, making relative ries of consumer judgment and decision-making. For one, it high- attribute judgments) can make consumers less confident in their at- lights the importance of a kind of confidence that has not previously tribute knowledge. This reduced confidence leads to more moderate been studied, namely, confidence in one’s own understanding of (less extreme) product evaluations and affects the options consumers common product attributes. This research demonstrates that changes choose. The authors derive an explanation for this phenomenon by in this kind of confidence can change consumers’ product evalua- drawing on research on folk science and the illusion of explanatory tions. depth. This research also demonstrates the counterintuitive finding that Previous research in the domain of consumer confidence is di- several common consumer activities frequently involved in making verse in scope, but has generally focused on confidence in the out- decisions—making estimates, making relative comparisons, explain- come of some judgment or evaluation (Brenner et al. 1996; Griffin ing an attribute—are not inert, but can change the subsequent evalua- and Tversky 1992; Karmarkar and Tormala 2010). This paper inves- tions and decisions the consumer makes. This research suggests that tigates a novel type of confidence: the implicit confidence consumers anything in the decision-making process that encourages consumers feel in their understanding of common product attributes. The ability to reflect on their attribute understanding could lead to less extreme to translate objective attribute information into subjective evalua- product evaluations and thereby affect choice. tions is fundamental to theories of consumer choice (Helson 1964; Although previous research has examined related issues such Janiszewski and Lichtenstein 1999; Kahneman and Tversky 1979; as how uncertainty can lead to purchase delay (Dhar 1997; Dhar and Sharif and Hovland 1961), but consumers’ confidence in their under- Simonson 2003) and how familiarity and confidence can facilitate standing of the attributes they use to make these evaluations has not judgments (Karmarkar and Tormala 2010; Park and Lessig 1981), it been investigated. We draw on insights from the literature in cogni- has not specified how the process of evaluating commonly encoun- tive psychology on folk science and the illusion of explanatory depth tered information such as product attributes can lead to more biased (Keil 2003; Rozenblit and Keil 2002; Sloman and Fernbach 2016) judgments. We contribute to the literature by specifying and testing a to make the prediction that a variety of cognitive tasks—likely to be theoretical account based on overconfidence and uncertainty, which common during typical consumer decisions—can reduce consumers’ predicts that when consumers are prompted to re-assess their knowl- confidence in their own ability to make evaluations. As we show, this edge of a product attribute, they often become less certain of that leads to more moderate (less extreme) product evaluations and can knowledge. affect choice. We suggest that consumers are often more confident in their REFERENCES ability to evaluate common product attributes than is justified by Brenner, Lyle A., Derek J. Koehler, Varda Liberman, and Amos their actual knowledge. We further suggest that this implicit confi- Tversky (1996), “Overconfidence in Probability and Frequency dence may be easily shaken by a number of simple interventions that Judgments: A Critical Examination,” Organizational Behavior invite consumers to consider the depth of their knowledge of a given and Human Decision Processes, 65 (3), 212-219. attribute. We propose that the effect of these knowledge probes is to Dhar, Ravi (1997), “Consumer Preference for a No-Choice Option,” reduce consumers’ confidence in their own ability to evaluate attri- Journal of Consumer Research, 24 (2), 215-231. butes, leading to less extreme evaluations of that attribute, relative to ——— and Itamar Simonson (2003), “The Effect of Forced Choice an otherwise equivalent consumer with unshaken confidence. on Choice,” Journal of Marketing Research, 40 (2), 146-160. A series of six experiments demonstrate that knowledge probes Griffin, Dale and Amos Tversky (1992), “The Weighing of related to common product attributes can influence subsequent prod- Evidence and the Determinants of Confidence,”Cognitive uct evaluations. Specifically, participants’ subjective evaluations Psychology, 24 (3), 411-435. of attributes like gigabytes, watts, and calories, were less extreme Janiszewski, Chris and Donald R. Lichtenstein (1999), “A Range following a knowledge probe than were evaluations made without Theory Account of Price Perception,” Journal of Consumer one. The experimental evidence suggests that a knowledge probe re- Research, 25 (4), 353-368. lated to an attribute reduces consumers’ confidence in their ability to Kahneman, Daniel and Amos Tversky (1979), “Prospect Theory: evaluate that attribute. This account was supported by a mediation An Analysis of Decision under Risk,” Econometrica, 263-291. analysis (Experiment 3), as well as by theoretically derived moderators. Specifically, when participants had a high level of attribute

Karmarkar, Uma R. and Zakary L. Tormala (2010), “Believe Me, I Rozenblit, Leonid and Frank Keil (2002), “The Misunderstood Have No Idea What I’m Talking About: The Effects of Source Limits of Folk Science: An Illusion of Explanatory Depth,” Certainty on Consumer Involvement and Persuasion,” Journal Cognitive Science, 26, 521-562. of Consumer Research, 36 (6), 1033-1049. Sharif, Muzafer and Carl I. Hovland (1961), Social Judgment, New Keil, Frank C. (2003), “Folkscience: Coarse Interpretations of a Haven, CT: Yale University Press. Complex Reality,” Trends in Cognitive Science, 7 (8), 368- Sloman, Steven and Philip Fernbach (2016), The Knowledge 373. Illusion: Why We Never Think Alone, New York, NY: Park, C. Whan and V. Parker Lessig (1981), “Familiarity and its Riverhead Books. Impact on Consumer Decision Biases and Heuristics,” Journal of Consumer Research, 8 (2), 223-230. Gender Identity and Liminality: An Exploration among Young Women with Breast Cancer Kathrynn Pounders, University of Texas at Austin, USA Marlys Mason, Oklahoma State University, USA

EXTENDED ABSTRACT Textual data was analyzed using an iterative part-to-whole Individuals must consume and interact with socially construct- method of the constant comparison (Spiggle, 1994). Research- ed, cultural representations of the self and the body (Pavia and Ma- ers developed an understanding of the emic themes and concepts son 2004; Scaraboto and Fischer 2013; Thompson and Hirschman that emerged first in each interview and next across all interviews 1995; Thompson and Ustuner 2015). When consumers’ internalized (Thompson, Locander and Pollio 1989). In an ongoing, recursive views or outward performativity conflict with such representations, manner, the researchers consulted the literature to contextualize the identity may be threatened or questioned. This study explores the narratives within gender identity. challenges that young women with breast cancer experience as they A predominant theme that emerged among the informants was a navigate embodiment and the intersectionality of their illness and liminal gender identity. More specifically, informants shared experi- gender identity. encing alterations of the body, changes in their motherhood identity, Consumer bodies that don’t conform to normative representa- and constraints in their sexual identity. The alterations of the body tions of what a woman should look like or how she should act may dramatically transformed their gender identity, particularly in how face exclusion in the marketplace. Heteronormative gender discourse they view themselves as a “woman.” Treatment for breast cancer in suggests that the body is a significant constituent of gender identity, terms of chemotherapy and surgery led informants to experience tre- and that bodily schemas are significant in distinguishing between the mendous disparity and unfamiliarity with their body. This resulted in heteronormative binaries of masculinity and femininity (Bourdieu a shattered sense of their sexual identity as well as physical limitation 1986; 1997). Accordingly, gender discourses are significantly de- or inability to enact their role of being a mom. Thus, the disconnect fined in bodies through habituated behavioral tendencies and social of embodiment was so deeply experienced that it commonly shat- schemas (Illouz 2012). Embodied identities that transgress from het- tered their sense of self and gender identity. Further, the majority eronormative gender discourse may confront challenges in consump- of the informants reject this “new” version of herself. Although the tion and acceptance (Coskuner-Balli and Thompson 2013). While informants express reaching the actualization of liminality, they have consumers can contest identities through collective unifying actions not reached aggregation (Schouten 1991; van Gennep 1960). Rather, and resignifying practices (Scaraboto and Fischer 2013; Thompson they reject the revised self, which results in extreme dissonance and and Üstüner 2015), at the core of such acts exists a presumption of an emotional turmoil. This is consistent with prior work that has shown empowered consumer who embraces the embodied aspects of iden- periods of liminality are associated with shock, confusion, and dis- tity. However, this is not the case for consumers facing illness and orientation (McCracken 1997; Schouten 1991), or at times a dam- limitations emanating from the body. aged sense (Turner 1967). Following the diagnosis of a life-threatening or disabling ill- Another emergent theme was using specific consumption prac- ness, consumer identities are abruptly and profoundly disrupted (Pa- tices in marketplace performance to express female gender identity. via and Mason 2004). With breast cancer, the visible alterations that Turner (1967, 1969) argues that during liminality, persons become occur to the body (e.g., mastectomy, hair loss) affect core cultural structurally invisible (lost between and betwixt) and socially am- markers of gender (Young 2014), and thus have resounding reper- biguous. The use of specific products to enact gender in public was cussions for gender identity. Previous gender expressions, such as a common theme, and is consistent with literature that has exam- styling one’s hair or dressing to highlight a curvy body, are compro- ined interactions in the marketplace as social performances affiliated mised due to the effects of treatment on body markers of gender and with cultural roles and scripts (Üstüner and Thompson 2015). These femininity. The need for control of the body and appearance is well- women described wearing wigs, jewelry, and makeup during a lim- established (Thompson and Hirschman 1995). Thus, the context of inal period to manage the expression of female gender identity when breast cancer diagnosis and treatment represents a unique situation, embodiment could no longer serve this purpose. as women have little to no control over changes to their body. Our In sum, consumer research, studies have investigated bod- understanding of the identity challenges that women face and how ies that do not conform to normative consumption spaces, ideal- they manage these challenges in light of such an acute, abrupt disrup- ized body images, and cultural representations of gender. However, tion emanating from the body is limited. these have tended to involve a relatively empowered or emancipated Seventeen depth interviews were conducted with young wom- consumer who has come to terms with their embodiment. Our in- en who had been diagnosed with breast cancer within the last five formants provide unique insights into the under examined area of years. The phenomenological approach was followed for attaining liminality (Thomassen 2016; Turner 1967, 1969), and the reflexive an in-depth understanding of an individual’s lived experiences (Mc- realization that they no longer fit with cultural notions of gendered Cracken, 1989; Thompson et al., 1989). Informants ranged from 23- discourse and habitus (Bourdieu 2001). Findings from this study 41 years old and represented diagnosis stages 1-4. All women were show that the intersecting facets of embodiment, illness, and gender currently undergoing treatment, including chemotherapy, radiation, created profound rippling effects in identity which the women had and surgeries. The sample was diverse in terms of family status and not yet reconstructed. Future research should explore the paths and education, and included those currently employed and unemployed; means by which women alter, integrate and reconstruct identities that married, divorced and never married; with children and without. The express gender outside cultural representations interviews began with grand tour questions (McCracken 1988) about the informant’s cancer experience in general. Informants were then asked to discuss how the diagnosis and subsequent treatment impacted their identity.

Advances in Consumer Research 824 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 825 REFERENCES Thomassen, Bjorn (2016), Liminality and the Modern: Living Bourdieu, Pierre (1986), “The Forms of Capital,” In Handbook of Through the In-Between, NY: Routledge. Theory and Research for the Sociology of Education. Edited Thompson, Craig J. and Elizabeth C. Hirschman (1995), by: Richardson J.” “Understanding the Socialized Body: A Poststructuralist Bourdieu, Pierre (1997), “Bodily Knowledge,” in Pascalian Analysis of Consumers’ Self-Conceptions, Body Images, and Meditations, Stanford, CA: Stanford University Press: 128- Self-care Practices.” Journal of Consumer Research 22 (2), 163. 139-153. Coskuner-Balli, Gokcen, and Craig J. Thompson (2013), “The Thompson, Craig J., William B. Locander and Howard R. Pollio Status Costs of Subordinate Cultural Capital: At-home Fathers’ (1989), “Putting Consumer Experience Back into Consumer Collective Pursuit of Cultural Legitimacy through Capitalizing Research: The Philosophy and Method of Existential- Consumption Practices.” Journal of Consumer Research 40 Phenomenology,” Journal of Consumer Research, 16 (2), (1), 19-41. 133-146. Illouz, Eva (2012), Why Love Hurts: A Sociological Explanation, Thompson, Craig J. and Tuba Üstüner (2015), “Women Skating Malden, MA: Polity. on the Edge: Marketplace Performances as Ideological Leigh, Susan (1994), “Cancer Survivorship: A Consumer Edgework.” Journal of Consumer Research, 42 (2), 235-265. Movement,” Seminars in Oncology, 21 (6), 783-786. Turner, Victor (1967), “Betwixt and Between: The Liminal Period McCracken, Grant (1987). Culture and Consumption Among the in Rites de Passage.” In The Forest of Symbols: Aspects of Elderly: Three Research Objectives in an Emerging Field,” Ndembu Ritual, Ithaca, NY: Cornell University Press, 93-111. Aging and Society, 7(2), 203-224. Turner, Victor (1969), “Liminality and Communitas,” in The Ritual Pavia, Teresa M. and Marlys J. Mason (2004), “The Reflexive Process, Structure and Anti-Structure. Chicago: Aldine, 94- Relationship Between Consumer Behavior and Adaptive 130. Coping,” Journal of Consumer Research 31 (2), 441-454. van Gennep, Arnold (1960), The Rights of Passage (M. B. Vizedom Scaraboto, Diane and Eileen Fischer (2013), “Frustrated & G. L. Caffee, Trans.). Chicago: University of Chicago Press. Fatshionistas: An Institutional Theory Perspective on Young, Iris Marion (2014). “Breasted experience: The look and Consumer Quests for Greater Choice in Mainstream Markets.” the feeling.” In The Politics of Women’s Bodies: Sexuality, Journal of Consumer Research, 39 (6), 1234-1257. Appearance, and Behavior. Edited by Weitz, Rose, and Schouten, John W. (1991), “Selves in transition: Symbolic Samantha Kwan. New York: Oxford University Press, pgs. consumption in personal rites of passage and identity 107-119. reconstruction.” Journal of Consumer Research 17(4), 412- 425. Spiggle, Susan. “Analysis and Interpretation of Qualitative Data in Consumer Research.” Journal of Consumer Research 21 (3), 491-503. Examining the Effect of Gender on Thinking Style, Price-Quality Perceptions, and the Actual Prices that Consumers Pay Vahid Rahmani, Rowan University, USA Elika Kordrostami, Humboldt State University, USA John Ford, Old Dominion University, USA

EXTENDED ABSTRACT pants were randomly assigned to one of two conditions in which they Consumers’ generally associate higher prices with higher levels reviewed the information of three brands of digital camera. One of of quality (e.g., Dawar and Parker 1994). Nevertheless, the relationship them was the target brand, Rumax, and the other two (i.e., Nikon and between price and objective quality in the marketplace is not always Canon) provided baseline price information. The presented informa- strong or even positive (e.g., Boyle and Lathrop 2009). This could be tion in both conditions was identical except for the price of the target explained by either consumers’ lack of access to product information camera which was either $80 or $130. After reviewing the brand in- (which is unlikely as we live in the age of information) or their reluc- formation, participants were asked to examine two photos ostensibly tance/inability to assimilate the available information to modify their taken by Rumax. Findings showed that price created a placebo effect price-quality judgments (ELM; Petty and Cacioppo 1986). The current in men and they had a more positive judgment of the pictures that they research is built on this latter assumption and attempts to answer the thought were taken by a more expensive camera. Whereas, the effect following question: What is the effect of gender on consumers’ price- of price on women’s judgment was not statistically significant. quality perceptions? In study 5a-5e, we collected the available information of 4200 Selectivity Model (Meyers-Levy and Maheswaran 1991) sug- athletic shoes, 3800 formal shoes, 781 suits, 909 jackets, and 597 gests that women (men) have a systematic (schematic) thinking style, socks, from Amazon.com. Findings of studies 5a-5e alongside the and they have a higher (lower) tendency to search for more details in findings of 5 pretest studies suggested that regardless of the product the decision-making process. Therefore, it could be posited that men type, men always paid a higher price to buy comparable products than should be more likely to use price-quality heuristics to make product women did. The price premiums that men paid ranged from 17% in the judgments; whereas, women should be more prone to make their qual- socks product category to 84% in the formal shoes product category. ity judgments based on a wider range of available information. Fur- Furthermore, findings showed that the observed price difference in thermore, if the price has a stronger effect on men’s than on women’s men’s versus women’s product categories was not attributable to con- perception of quality, the average prices of men’s products should be sumer involvement, shopping frequency, type of product, manufactur- higher than the average prices of comparable women’s products. ing cost, online review information, available product information, and In nine studies, which employed survey research, experimental brand name. research, and observational research methods to achieve methodological triangulation, we investigated the effect of gender on consumers’ General Discussion motivation/ability to process the information (studies 1 and 2), their Current study makes several substantive contributions to theory price-quality perceptions (studies 3 and 4), and the actual prices that and practice. First, to our knowledge, this paper is the first to show that they pay for similar products (studies 5a-5e). Next, the findings of gender influences consumers’ price-quality perceptions (studies 3 and these studies are briefly reported next. 4). Second, this paper provides further empirical support for selectivity Study 1 was designed to measure the effect of gender on consum- model (studies 1 and 2). Third, this research contributes to the adver- ers’ ability to pay attention to readily available information and form tising literature by showing that gender affects the target audience’s accurate judgments in the context of the print advertisement. Respon- attitude towards the deceptive advertisement (study 1). Fourth, this dents were randomly assigned to one of two advertisement conditions, paper contributes to the digital marketing literature by showing that one of which had a manipulative ad claim. Results showed that type of the effect of online review valence is greater than the effect of online ad affected women’s attitude towards advertisement. But men did not review volume on a product’s price (studies 5a-5e). Finally, this paper notice the suspicious claim in the manipulative advertisement. offers important practical implications. Findings suggest that opposite Study 2 was designed to investigate whether gender influences pricing strategies could be used for men and women product categories consumers’ ability to make accurate judgments of the product qual- to increase the profitability (studies 5a-5e). ity based on the available product information. Respondents were randomly assigned to one of two conditions that entailed a table of REFERENCES information about 25 brands of digital camera. Next, they evaluated Boyle, Peter J. and E. Scott Lathrop (2009), “Are Consumers’ the quality of ten fictitious models of a digital camera based on the Perceptions of Price–Quality Relationships Well Calibrated?” information that they had reviewed. Findings suggested that women’s International Journal of Consumer Studies, 33(1), 58-63. judgments of the relationship among product attributes corresponded Dawar, Niraj and Philip Parker (1994), “Marketing Universals: with the information that they had reviewed; whereas men’s judgments Consumers’ Use of Brand Name, Price, Physical Appearance, did not reflect the presented information in the tables. and Retailer,” Journal of Marketing, 58(2), 81-95. Study 3 was designed to provide an initial examination of the Meyers-Levy, Joan and Durairaj Maheswaran (1991), “Exploring potential effect of gender on the general-price-quality perceptions. Differences in Males’ and Females’ Processing Strategies,” Respondents answered several questions in a survey. Results showed Journal of Consumer Research, 63-70. that the price-quality perception was significantly stronger in men than Petty, Richard E., and John T. Cacioppo (1986), “The Elaboration in women. Furthermore, the average price-quality perception score Likelihood Model of Persuasion,” Communication and was significantly above the midpoint value of 4.0 for men, but not for Persuasion, 1-24. women. Study 4 was designed to establish the generalizability of the findings of study 3 by employing a more rigorous research method. Partici-

Advances in Consumer Research 826 Volume 45, ©2017 The Begging Game: On the Power of the Ask in Charitable Exchange Mark Ratchford, Tulane University, USA Ethan Pew, Stony Brook University, USA Bart Victor, Vanderbilt University, USA

EXTENDED ABSTRACT More than half (52.5%) of “askers” requested $5 of the “owner’s” Much of charitable giving results from formal solicitations. endowment. Those asks were accepted 61.9% of the time. Asking However, asking has been largely ignored in the altruism and gen- for $4 was accepted 75% of the time, and asking for $3 was accepted erosity literature, which relies heavily on the dictator and ultimatum 80% of the time. In the DG conditions, dictators transferred $3.50 games. Through the begging game (a variant of the dictator game) (SD=2.79) to the passive recipient, while keeping $6.50 of the $10 we offer insights on the role asking plays in charitable exchange. sum. This result is in line with the findings of prior DG experiments An extensive literature in the social sciences describes a system- (e.g., Engel, 2011; Hoffman, McCabe, and Smith, 1996), albeit atic variation from rational self-interest in which relatively resource slightly more generous in giving than in prior studies. The more gen- rich individuals have been shown to give more than absolutely re- erous DG outcome is largely the result of three (of 40) “dictators” quired to those with fewer resources. For example, in the two-player who transferred the entire $10 sum. dictator game (Kahneman, Knetsch, and Thaler, 1986; Forsythe et Study 2 (N=404) was conducted using an online panel. Partici- al., 1994), a proposer (dictator) offers a split of an endowment to pants completed this online study in exchange for reward-contingent a recipient who is entirely passive. The recipient gets whatever the monetary compensation of between $0 and $10. The average age of proposer offers to them and cannot reject or counter-propose the en- respondents was 51.9 (range 18 to 84), 52% were female, average dowment. A purely rational dictator should propose nothing to the income was just over $60,000 per year and 84.8% had at least some recipient. However, findings typically suggest the majority allocate college education. Subjects were randomly assigned to one of two at least some money to the passive recipient. Only about 40% keep conditions, either owner or asker in the BG, with approximately 70% the entire sum, while the average amount given (by the other 60%) of incoming subjects randomly assigned to the “owner” condition. is about 20% of the sum (Guala and Mittone, 2010), with modal al- All participants received the same instructions and overview for the locations at 50% and zero (Fehr and Schmidt, 2006). Researchers BG as in Study 1. However, for this online study participants played have attempted to explain this tendency toward “irrational” giving against a computer with a pre-programed decision strategy. On aver- by focusing on social and psychological influences such altruism, age askers requested $4.71 (SD=1.59) of the owner’s $10, with no warm glow giving, social pressure (e.g., experimental observation), significant effects of age, gender, income, or other demographics. In or some combination of those motivations (e.g., Andreoni, 1989, all, 51.3% of askers requested $5, while another 28.3% asked for 1990; DellaVigna, List, and Malmendier, 2012; Schmitz, 2016). This $4. For those assigned to the owner condition, requests of $3 or less stream of research has in turn spawned numerous normative models tended to be accepted (73.8% overall acceptance rate for offers in this of the impact of “social preferences,” “interdependent preferences,” range). In comparison, requests of $4 and $5 were substantially more and “intention based reciprocity” on individual utility functions likely to be rejected (42.3% acceptance rate for offers in this range). (Fehr and Schmidt, 2006). Study 3 (N=80) was conducted at a large university in the Unit- In the present paper, we study the “power of the ask,” and find ed States. Undergraduate students took part in this study in exchange the mere act of asking for money shifts power to the solicitor. The for both course credit and reward-contingent monetary compensa- findings across three studies suggest that while giving in response to tion of between $0 and $10. The protocols were identical to the BG a direct solicitation may in part be a function of the altruism of the condition in Study 1 with the exception that owners could transfer giver, the ask request itself may prime an aversion to violating an any amount to the asker. They were not bound by the binary “accept” implicit norm of fairness (Forsythe et al., 1994; Hoffman, McCabe, or “reject” options that appeared in Study 1. In this study askers re- and Smith, 1996). Asking imposes a norm of equitable exchange via quested an average of $6.48 (SD=2.48), which is significantly greater the amount requested so long as the request is not unreasonably high than the average ask in Study 1 (M1=5.33; t(78)=2.2, p=.03) and (more than half in our studies). We contend that violating this norm Study 2 (M2=4.71; t(151)=5.17, p<.001). The average amount trans- of equitable exchange bestows discomfort upon the recipient of the ferred was $3.03 (SD=1.98), which is directionally higher than the solicitation via felt social pressure. Hence the solicitation or “ask” average of transfer in Study 1, but not significantly so (t(78)=1.65, can alter the nature of the exchange. Overall, this suggests that altru- p=.10). ism may be a subordinate driver of economic transfers in the dictator and ultimatum games. REFERENCES Study 1 (N=160) was conducted at a large university in the Andreoni, J. (1989). Giving with impure altruism: Applications United States. Undergraduate students took part in this study in ex- to charity and Ricardian equivalence. Journal of political change for both course credit and reward-contingent monetary com- Economy, 97(6), 1447-1458. pensation of between $0 and $10. Subjects were assigned to play Andreoni, J. (1990). Impure altruism and donations to public either the begging game (BG) or the dictator game (DG), with 50% goods: A theory of warm-glow giving. The economic journal, of incoming subjects randomly assigned to each. The DG followed 100(401), 464-477. the standard protocol. In the begging game, the player who did not DellaVigna, S., List, J. A., & Malmendier, U. (2012). Testing for receive money (the asker) was provided a chance to ask the player altruism and social pressure in charitable giving. The quarterly who did receive funds (the owner) for a portion of that money. Own- journal of economics, 127(1), 1-56. ers could then accept or reject the ask. On average askers requested Engel, C. (2011). Dictator games: A meta study. Experimental $5.33 (SD=2.18) of the owner’s $10, and owners accepted 52.5% of Economics, 14(4), 583-610. asker’s requests overall. As expected, whether or not an offer was accepted was a function of the asked for amount (χ2(7)=13.93, p=.05).

Fehr, E., & Schmidt, K. M. (2006). The economics of fairness, Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1986). Fairness reciprocity and altruism – experimental evidence and new and the assumptions of economics. Journal of business, theories. Handbook of the economics of giving, altruism and S285-S300. reciprocity, 1, 615-691. Schmitz, J. (2016). Is Charitable Giving a Zero Sum Game? An Hoffman, E., McCabe, K., & Smith, V. L. (1996). Social distance Experimental Investigation of Why People Give to Charity. and other-regarding behavior in dictator games. The American (November 1, 2016). Available at SSRN: https://ssrn.com/ Economic Review, 86(3), 653-660. abstract=2862479 I’m Not Your Friend, Buddy: Friendliness Without Friendship Hinders Consumer-Firm Relationship Development Suzanne Rath, Queen’s University, Canada Laurence Ashworth, Queen’s University, Canada Nicole Robitaille, Queen’s University, Canada Matthew Philp, Universite de Montréal, Canada

EXTENDED ABSTRACT and a (marginally) significant interaction suggested that the negative Although the literature on consumer-brand interactions has effect of friendliness was greater in desirable stores than in undesir- identified a variety of different ways in which firms can forge rela- able stores. Although somewhat surprising at first glance, examina- tionships with their customers, one of the central, practical sugges- tion of the means indicated the desire to stay in the undesirable store tions is to treat consumers with friendship. Several consistent char- was universally low and that the nature of the interaction did little to acteristics of friendship have been identified by previous research change that. In contrast, in a desirable store, friendliness reduced an to help define the friendship role, including “opening up” and shar- otherwise strong desire to stay in the store. The final study examined ing personal information (Price & Arnould, 1999; Grayson, 2007). whether friendliness was really most appropriate in the context of Friendship is one of the most common forms of relationship (Chan & a pre-existing friendship, rather than at the beginning of a possible Cheng 2004; Hartup 1975; Hays 1985, 1988) and seems highly rel- relationship. We found a negative effect of friendliness on desire to evant to the context of consumer-brand and consumer-firm relation- stay when participants had no prior relationship with the salesperson, ships (e.g., Aggarwal 2004). Being friendly and engaging in friend- and in contrast, a positive effect of friendliness when participants like behaviors is commonly prescribed in the practitioner literature had a prior relationship with the salesperson. These results were con- (e.g., Cummings 2015; Martin 2010; Turnali 2014) and in textbooks sistent with the idea that friendly behaviours are a consequence of (Baron 1997). Moreover, a number of studies support the idea that friendship, rather than a cause of it. Therefore, it seems that friendli- friendliness, at least in a service context, leads to higher levels of ness in the absence of friendship appears to backfire. The results of satisfaction (Brown & Sulzer-Azaroff 1994; Pugh 2001; Tsai 2001; these four studies demonstrate a consistent pattern that when firms Tsai & Huang 2002; Barger & Grandey 2006). act like friends, consumers react less favorably compared to when While friendly behaviors are clearly symptomatic of friendship, the firm acts neutrally. it is less clear that they are causally responsible for friendship or that The current research contributes to the consumer-brand rela- they are even appropriate in the early stages of friendship (i.e., when tionship literature by demonstrating that acting like a friend at firm there is no friendship). Studies of relationship development, for ex- touchpoints may not be the best way for firms to develop a relation- ample, show that relationship “initiation” behaviors are very different ship with consumers. While the literature seems to imply that con- than latter-stage behaviors (Fox, Warber, & Makstaller 2013; Knapp sumers view firms and brands as meaningful relationship partners, 1978). Other work has shown that behaviors that conflict with the our studies suggest that, in fact, consumers may not characterize their expectations for the relationship, such as whether it is business or relationships with firms and brands in this way. In four studies, we friendship (Grayson, 2007), can elicit discomfort or unease (Clark & demonstrate a robust, counter-intuitive effect: acting like a friend Mills, 1993) or even undermine relationships entirely (Fletcher et al. can hinder relationship development efforts between firms and con- 2000). Thus, to the extent that friendly behaviors are not character- sumers. Evidence from four studies demonstrates that friendly en- istic of the behaviors people generally engage in at the beginning of counters reduce consumers’ perceived comfort such that they would a relationship, it seems possible, contrary to common wisdom, that rather not stay or continue the interaction, thus impeding the relation- friendly behaviors may cause discomfort and might even harm the ship development process. So, acting like a friend reflects a customer firm’s chances of establishing a friendship. service strategy that may unintentionally do more harm than good. We conducted four studies to test various elements of our prediction. Study 1 tested the basic idea that consumers may react nega- REFERENCES tively to overtly friendly behavior by a firm representative. Friendly Aggarwal, Pankaj (2004), “The Effects of Brand Relationship vs. neutral scenarios for each study were designed based on the con- Norms on Consumer Attitudes and Behaviour,” Journal of ceptualization of friendly behaviours described by prior research Consumer Research, 31 (June), 87-101. (Price & Arnould, 1999; Grayson, 2007). We found initial evidence Barger, Patricia. B. and Alicia A. Grandey (2006), “Service with that friendly firm encounters were less favorable to consumers than a Smile and Encounter Satisfaction: Emotional Contagion neutral firm encounters. Studies 2 and 3 examined possible boundary and Appraisal Mechanisms,” The Academy of Management conditions to the negative effect of friendliness documented in Study Journal, 49 (6), 1229-1238. 1, and statistically ruled out perceived ulterior motives as a possi- Baron, Gerald. R (1997), “Friendship Marketing: Growing Your ble alternative explanation. Specifically, study 2 examined whether Business by Cultivating Strategic Relationships,”Grants Pass, the negative effect might be mitigated or reversed when consumers OR: The Oasis Press. could anticipate a future relationship with the firm. Not only did the Brown, Carolyn. S. and Beth Sulzer-Azaroff (1994), “An negative effect found in study one replicate, we also found that the Assessment of the Relationship Between Customer observed negative effect of friendliness was not simply an artifact of Satisfaction and Service Friendliness,” Journal of our manipulation – we saw a positive effect of friendliness in a con- Organizational Behaviour Management, 14 (2), 55-75. text where friendliness should have been normative. Also, we found Chan, Darius. K. S. and Grand H. L. Cheng (2004), “A Comparison the same negative effect of friendliness in the context where consum- of Offline and Online Friendship Qualities at Different ers might anticipate a future relationship. Study 3 examined whether Stages of Relationship Development,” Journal of Social and the effect was moderated by the desirability of the firm. We found Personality Relationships, 21, 305-320. that the negative effect found in the previous two studies replicated,

Clark, Margaret S. and Judson Mills (1993), “The Difference Knapp, Mark. L (1978), “Social Intercourse: From Greeting to Between Communal and Exchange Relationships: What it is Goodbye,” Needham Heights, MA, USA: Allyn & Bacon. and is Not,” Personality and Social Psychology Bulletin, 19 Martin, Steve. W (2010, October 13), “Persuasion Tactics of (6), 684-691. Effective Salespeople,” Retrieved September 11, 2016, from Cummings, Carrie (2015, October 11), “Infographic: Brands https://hbr.org/2010/10/persuasion-tactics-of-effectiv Should Treat Consumers as Friends, Not Age Groups,” Price, Linda. L. and Eric J. Arnould (1999), “Commercial Retrieved September 11, 2016, from http://www.adweek.com/ Friendships: Service Provider-Client Relationships in Context. brandmarketing/infographic-brands-should-treat-consumers Journal of Marketing, 63(4), 38-56. friends-not-age-groups-167492/ Pugh, S. Douglas (2001), “Service with a Smile: Emotional Fletcher, Garth. J. O., Jeffrey A. Simpson, and Geoff Thomas Contagion in the Service Encounter,” Academy of (2000), “Ideals, Perceptions, and Evaluations in Early Management Journal, 44 (5), 1018-1027. Relationship Development,” Journal of Personality and Social Solomon, Michael. R, Carol Surprenant, John A. Czepiel, and Psychology, 79 (6), 933-940. Evelyn G. Gutman (1985), “A Role Theory Perspective on Fox, Jesse, Katie M. Warber, and Dana C. Makstaller (2013), “The Dyadic Interactions: The Service Encounter,” Journal of Role of Facebook in Romantic Relationship Development: An Marketing, 49 (winter), 99-111. Exploration of Knapp’s Relational Stage Model,” Journal of Tsai, Wei-Chi (2001), “Determinants and consequences of Social and Personal Relationships, 30 (6), 771–794. employee displayed positive emotions,” Journal of Grayson, Kent (2007), “Friendship versus Business in Marketing Management, 27 (4), 497-512. Relationships,” Journal of Marketing, 71 (4), 121-139. Tsai, Wei-Chi, and Yin-Mei Huang (2002), “Mechanisms linking Hartup, W. W (1975), “The Origins of Friendships,” In Friendship employee affective delivery and customer behavioural and Peer Relations, ed. M. Lewis and L. A. Rosenblum, New intentions,” Journal of Applied Psychology, 87 (2), 1001-1008. York, NY: John Wiley & Sons, 11-26. Turnali, Kaan (2014, December 14), “4 Reasons Why Excellent Hays, Robert. B (1988), “Friendship,” In Handbook of Personal Customer Service Should Start With A Smile,” Retrieved Relationships, ed. S.W. Duck, New York: John Wiley & Sons, September 11, 2016, from https://www.forbes.com/sites/ 391-408. sap/2014/12/14/4-reasons-why-excellent-customer-service Hays, Robert. B (1985), “A Longitudinal Study of Friendship should-start-with-a-smile/#6980746129db Development,” Journal of Personality and Social Psychology, 48 (4), 909-924. Customers’ Magnitude Perception for Letters Versus Numbers Shelly Rathee, University of Utah, USA Arul Mishra, University of Utah, USA Himanshu Mishra, University of Utah, USA

EXTENDED ABSTRACT bets because people learned them that way as children (e.g., A-B-C, Prior research in marketing has shown that cues do not need to L-M-N-O-P), influencing how letters come to mind more fluently. be diagnostic, and over time relying on cues becomes so habitual that Therefore, people use the fluency with which letters come to mind people do so even without awareness (Payne, Bettman, and Johnson (arising from subgrouping structures) as a heuristic to estimate mag- 1991, 1994). Although prior work has not provided explicit distinc- nitude. The heuristic is that if letters flow fluently in their minds, they tions, cues encountered in everyday life can have quantitative or believe that the task using the letter cues would be similarly fluent, qualitative features. and thus the magnitude required to perform the task would be less. Quantitative cues embed numeric information (Thomas and Importantly, we predict that such a pattern will reverse in the Morwitz 2005; Monga and Bagchi 2012) with inherent cardinal performance stage because individuals translate the fluency with let- properties (e.g., numbers), whereas qualitative cues embed verbal in- ters into an expectation of low magnitude estimates. However, on formation with ordinal properties (Svenson and Karlsson 1986). For actually performing the act using letters that contain little magnitude example, attribute information can be communicated in terms of nu- information, they will experience a violation of their expectations of meric information such as volume, density, or price. Estimating mag- ease and perceive that they are expending greater effort and taking nitude from quantitative cues is easy because of their cardinal prop- more time performing an action than if they were using number cues. erties (e.g., 250 lbs is 100 lbs more than 150 lbs). However, on many occasions people need to infer magnitude from qualitative cues, such Data Analysis & Results as the terror/safety levels ranging from green to red, stores using Studies 1a–2 tested our theoretical predictions in the estimation colors to indicate crowd density, or the distance to walk between stage, study 3 the performance stage, and study 4 tested both stages. airport exit gates numbered with letters. In this research, we focus We designed Studies 1a and 1b to demonstrate the proposed ef- on numbers (1, 2, 3…) to characterize quantitative cues and letters fect in a retail store and seating domain setting respectively. In study (A, B, C…) to characterize qualitative cues. Both letters and numbers 1a – participants were shown a store picture with five aisles labeled are ordered left to right, are single digit, and are commonly used as 1 to 5 (A to E) and were asked to imagine scenario in which they as labels. We examine whether providing letters versus numbers to were shopping in aisle 8 (or H) and realized that they had forgotten to estimate magnitude, such as distance or time, can produce differing pick up a product in aisle 1 (or A). We then asked them, “Would you outcomes for consumers. Consider, for example, a person exercis- be willing to walk back to the aisle to pick up the product, considering on a treadmill-using letters (e.g., “completed level A, B, C …”) ing that you could collect the item on the next visit?” Participants rather than numbers (e.g., “completed level 1, 2, 3. …”), appearing responded on a seven-point scale (1 = highly unlikely, 7 = highly on the monitor as milestones to evaluate the exercise level. Would a likely). Consistent with our theorizing that individuals estimate less number cue lead to greater perceived and actual effort? Would there magnitude with letters, we found that participants were more will- be a difference between estimates before and after performing the ing to go back for the product they had forgotten if the aisles were action? The aim of this research is to answer these questions in ev- labeled with letters than with numbers (Mletter = 5.90 versus. Mnumber = eryday consumer contexts. 4.35; F(1, 72) = 14.843, p < .01, η2 = .170). Study 1b showed used seat-selection domain, whereby consum- Theoretical Background ers evaluate seats of a flight from the seating arrangement chart. As We focus on findings that letters and numbers will affect judg- part of the cover story, we showed participants the seating-plan of ments differently depending on whether people are contemplating an airplane and asked the participants to provide subjective distance taking an action or are actually performing the action. For ease of and effort (1 = far too little, 7 = far too much) estimates they thought reference, we refer to these two stages as the estimation stage and it would take them to walk to designated seats. Further, to measure the performance stage. When people are contemplating an action fluency, we asked individuals how easy or difficult was it to make with a letter cue, they are likely to generate a lower estimate (i.e., seat-selection with letters versus numbers. As expected, we found effort and time) than when they make the decision with a number that those in the letter cue condition provided a lower estimate than cue. However, a counter-intuitive reversal occurs when people per- those in the number (Mletter = 3.01 vs. Mnumber = 3.39; t(200) = -3.36, form an action. They perceive expending more effort with letters p < 0.001). Further, the results of the mediation analysis confirmed than with numbers. We predict differences between the estimation the mediating role of fluency in the relationship between the type of and performance stages using prior work on alphabet subgrouping cues and magnitude estimates (β = -0.032; 95% CI [-0.069, -0.006]). (Jou 2003; Fulbright et al. 2003) and expectation violation (Tzur and Study 2 manipulated fluency through providing broken sub- Berger 2008). groups of letters that reduced the reliance on fluent subgroup in let- Regarding our theoretical mechanism, we propose that when ters; this subsequently reduced the difference in magnitude estimates contemplating an action (estimation stage), individuals are likely to between letters and numbers. This paper-and-pencil study was a 2 estimate less magnitude with letters because such cues do not have (fluency: fluent vs. non-fluent) × 2 (cues: letters vs. numbers) design. clear magnitude information. Unlike numbers that possess both or- Participants were asked to place a mark on a 120-millimeter line, dinal and cardinal properties, letters possess only ordinal properties indicating how much distance they believed they would have walked (Gevers, Reynvot, and Fias 2003; Kara, Gunasti, and Ross 2015). from the start to the target row in the theater. The results of Study 2 Therefore, the lack of clear magnitude information from letters influ- showed that fluency played an important role in magnitude under- ence people to depend on subgrouping properties of letters. Work on estimation with letter cues compared with number cues (Mnumber/fluent letter cognition suggests that there are subgroups of letters in alpha- = 55.15 mms vs. Mnumber/non-fluent = 55.71 mms; t(96) = 0.1, p = 0.91;

Mletter/fluent = 41.74 mms vs. Mletter/non-fluent = 55.07 mms; t(94) = 2.61, p REFERENCES < 0.01). Fulbright, Robert K, Manson, Stephanie C, Skudlarski, Pawel, We conducted Study 3 in a local gym to measure people’s per- Lacadie, Cheryl M, and Gore, John C. (2003). “Quantity ceptions of effort expended depending on whether letters versus determination and the distance effect with letters, numbers, numbers marked the laps completed on a treadmill. It was demon- and shapes: a functional MR imaging study of number strated that in performance stage individuals who exercised using processing.” American journal of neuroradiology, letters perceived that they expended greater effort compared to those 24(February), 193-200. exercising using numbers (Mletter = 2.31, Mnumber = 1.62, F(1, 134) = Gevers, W., Reynvoet, B., & Fias, W. (2003). The mental 8.531, p = .004, η2 = .06). representation of ordinal sequences is spatially organized. Finally, in Study 4 we examined both estimation and perfor- Cognition, 87(March), B87-B95. mance stages for the same individuals by using process evidence and Jou, Jerwen. (2003). “Multiple number and letter comparison: video recording them as they walked from a start to an end location. Directionality and accessibility in numeric and alphabetic Results revealed that in the estimation stage, participants in the num- memories.” The American journal of psychology. ber cue condition estimated a greater magnitude than those in letter Kara, S., Gunasti, K., & Ross, W. T. (2015). Is it the ‘alpha’or the cue condition (Mnumber = 0.21 vs. Mletter = -0.20; F(123) = 8.27, p = ‘numeric’?: Consumers’ evaluation of letter versus number 0.005 , η2 = 0.06). We obtained process evidence for our feedback- changes in alphanumeric brand names. Journal of Brand related negativity hypothesis and demonstrated using walk-time that Management, 22(June), 515-533. individual using letter (versus number) cues slowed down while per- Monga, A., & Bagchi, R. (2012). Years, months, and days versus 1, forming the task (Tletter =14.09 seconds vs. Tnumber = 11.34 seconds, 12, and 365: The influence of units versus numbers.Journal of F(1, 122) = 17.57, p < .0001). Furthermore, we demonstrated the Consumer Research, 39(January), 185-198. downstream influences on marketing variables such purchase post- Payne, J., Bettman, J. R., & Johnson, E. J. (1991). Consumer ponement, impulsive choice, and preference for hedonic products. decision making. Handbook of consumer behaviour, 50-84. ____ (1994). The adaptive decision maker. Journal of the Conclusion Operational Research Society, 45(July), 850-850. In this research we demonstrate that qualitative versus quan- Svenson, O., & Karlsson, G. (1986). Attractiveness of decision titative cues (letters and numbers) produce differing estimates (of alternatives characterized by numerical and non‐numerical time, distance and effort). Moreover, we suggest that individuals un- information. Scandinavian Journal of Psychology, derestimate magnitude with letter cues in the estimation stage but 27(January), 74-84. take longer to perform the task with letter cues. Across three labora- Thomas, M., & Morwitz, V. (2005). Penny wise and pound foolish: tory experiments and two-field studies, we test our predictions in the left-digit effect in price cognition.Journal of Consumer domains such as health, shopping, and seat-selection. We also docu- Research, 32(January), 54-64. ment downstream effects of effort estimation with number and letter Tzur, G., & Berger, A. (2007). When things look wrong: Theta cues on purchase postponement, impulsive choice, and preference activity in rule violation. Neuropsychologia, 45(December), for hedonic products. 3122-3126. Ritualistic Meal Consumption and the Temporal Reduction of Divisions Due to Economic Disparity Ateeq Abdul Rauf, Information Technology University, Pakistan Ajnesh Prasad, Tecnologico de Monterrey, Mexico

EXTENDED ABSTRACT Data Collection In this paper, we illuminate how a social consumption practice The data for this study was collected from 2012-2016. As part in an ephemeral religious organization subverts systems of social of a more formal ethnography, the first author then undertook a 40- inequality that otherwise prevail in and structure society. Specifi- day sojourn with 10 other TJ participants. After the sojourn, pur- cally, drawing on a rich ethnographic study in Pakistan, we show posive sampling was used to choose an additional 13 informants to how sharing food in the Tablighi Jamaat (TJ)—a religious organi- interview to achieve better triangulation (Ritchie et al. 2013). zation originating in South Asia that is practiced intermittently by its followers—represents temporal spaces of egalitarianism. Within Findings these temporal spaces, entrenched social hierarchies that are salient From the data, we found that food consumption in TJ sojourns in organizing Pakistani society are deconstructed. We found that al- subverts the everyday social and economic hierarchies present in though the fundamental principles of the Tablighi Jamaat advocate Pakistani life. Our analysis yields two thematic explanations be- for transgression from the social hierarchies that propagate myriad hind this temporary collapse in social boundaries: changing causes inequalities by demarcating local Muslims into spheres of different (religious cause versus the everyday routine of earning to sustain social and economic classes, it is in the practice of food consumption and consume); and bringing everyone to the same consumption life- when challenges to these hierarchies become the most conspicuous. style. While these themes are exposed during TJ activities, they are With burgeoning social inequalities—which are only consecrat- temporal in their effect to linger on post-TJ and differ by the socio- ed by growing income and wealth disparities—we respond to recent economic profiles of the participants and their connection with the calls from scholars in the field to identify trajectories by which to movement. disrupt systems that create, maintain, and reify stratification between individuals in a society (Fotaki and Prasad 2015). We demonstrate Discussion how a seemingly inconsequential social practice—in this case, meal From the results of this study, we can derive some contribu- sharing (Belk 2010)—can function as a powerful mechanism by tory notes for existing literature. First, food consumption can be a which to destabilize entrenched social hierarchies that have histori- social bridging activity. It has been previously established that where cally structured the society in which the consumption practice oc- you eat, how you eat, and with whom you eat signifies social power curs. Notwithstanding this point, we do not intend to romanticize our or lack of it (Bourdieu 1984). Organizations in Pakistan like gov- argument. Indeed, we appreciate the fact that the temporal nature of ernment institutes, corporations, and schools (where teacher and the TJ and the ingrained nature of the cultural codification of social student separation is evident) use differing meal consumption prac- hierarchies in Pakistan should not be negated. We equally acknowl- tices (place, style, and group) to signal their social superiority. TJ edge that these realities of Pakistani society raise the question con- participation challenges the strongly embedded social stratification cerning the extent to which an apparently innocuous phenomenon prevalent in Pakistani society. Food consumption appears to be an such as food consumption transforms how individuals relate to one exemplar of this practice. another following the religious experience in spaces not governed by Second, changing ideological causes can help lessen social di- the TJ. While these remain fair points, we contend that, at the very vides. The foundation stone for driving the motivation of TJ travel- minimum, egalitarian food consumption practices in the TJ symboli- ers is their belief for success in a Hereafter which is accomplished cally reveal alternative tenets by which to organize individuals in through sacrificing and working for a higher spiritual cause in this society—principles based more on equality than on inequality. The world. We learn that a cause that is promoted to be larger than the present study extends prior literature by highlighting how social in- everyday routines of working to earn, providing and caring for the equality can be challenged through a consumption practice in a less- family, and consuming to live can be a means to displace the social industrialized country context. understandings that govern world citizens. In this regard, an egalitarian-motivated social cause, can aid in reducing barriers due to Methodology economic divides. This study uses multi-sited ethnography and in-depth interviews Third, while the concept of sharing has already been noted to be to understand how a consumption practice can collapse social bound- a means of bringing people closer (Belk 2010), this study illuminates aries that maintain systems of inequality. Specifically, we analyze that sharing can be a means of solving a cardinal social problem. food consumption by participants of the orthodox Islamic religious We additionally note that this study is limited in its focus to group Tablighi Jamaat (TJ). address economic disparity, while not considering other forms of inequality (religious, gender, etc.). Research Site We use the context of TJ to illustrate the phenomenon of food REFERENCES consumption as a normalizing social force. TJ forms a traditional re- Belk, Russell W. (2010), “Sharing,” Journal of Consumer Research, formist approach to Islam that developed in late colonial India. The 36 (February), 715–34. movement follows an approach accessible to the common Muslim, Bourdieu, Pierre (1984), Distinction : A Social Critique of the focusing on making participants lay preachers with periodic com- Judgement of Taste. Cambridge, MA: Harvard University mitments that involve staying away from their residences, families, Press. and workplaces. According to one estimate, TJ today has 80 million followers (Taylor 2009) in more than 200 countries.

Fotaki, Marianna and Ajnesh Prasad (2015), “Questioning Taylor, Jenny (2009), “What is the Tablighi Jamaat?” The Guardian, Neoliberal Capitalism and Economic Inequality in Business http://www.theguardian.com/commentisfree/belief/2009/ Schools,” Academy of Management Learning and Education, sep/08/religion-islam-tablighi-jamaat 14 (4), 556-575. Ritchie, Jane, Jane Lewis, Carol M. Nicholls, and Rachel Ormston (2013), Qualitative Research Practice: A Guide for Social Science Students and Researchers, London: Sage. Made by Mistake: When Mistakes Increase Product Preference Taly Reich, Yale University, USA Daniella Kupor, Boston University, USA Rosanna Smith, Yale University, USA

EXTENDED ABSTRACT Empirical Evidence Imagine a pastry chef is making a new batch of chocolate. Not Study 1 paying attention, the chef leaves the chocolate in the oven for an Study 1 examined the preference for products made by mistake extra five minutes by mistake, resulting in the chocolate having a dif- in a consequential choice setting. Participants learned about a new ferent taste than the chef had intended. Would learning of the chef’s chocolate that either was (or was not) the result of a mistake. They mistake increase or decrease consumer preference for the chocolate? were then given the choice of receiving either the chocolate or ex- Intuition and past research suggest that the knowledge of a mis- tra monetary compensation. A chi-square analysis revealed that the take (i.e., careless action) involved in the making of a product would chocolate made by mistake was chosen (88.6%) more often than the decrease consumer preference. Indeed, prior work reveals that mis- chocolate that was made intentionally (70.3%), χ 2(df = 2, N = 144) takes are perceived as undesirable and often result in negative infer- = 7.31, p = .007. ences about the person or company that made the mistake (Chesney and Su 2010; Michael 1976; Palmer, Simmons, and de Kervenoael Study 2 2010). Consequently, individuals and companies tend to avoid shar- Study 2 examined whether people still prefer a product made ing information about their mistakes with others (Edmondson 1996; by mistake even when the mistake detracts from a product (i.e., a Michael 1976; Stefaniak and Robertson 2010; Uribe et al. 2002). De- drawing of a face that is tarnished by a pen scribble). To examine this spite this previous literature, we propose that knowledge of a mistake prediction, participants viewed a drawing of a face with a scribble, in the creation of a product can ironically enhance consumer prefer- and either read information about the artist which included the fact ence for that product. We propose that this preference for products that the artist made a mistake while making the drawing (in the mis- “made by mistake” is not limited to mistakes that enhance a product, take condition), read information about the artist which included the but also extends to mistakes that detract from a product. fact that the artist did not make a mistake while making the drawing We posit that this preference results from how consumers rea- (in the no mistake condition), or did not read anything about the art- son about the relationship between intention and action. Specifically, ist (in the control condition). Thus, this design also examined our we draw from a large body of work on the intentionality bias (Begue theorizing that people still prefer products made by mistake even af- et al. 2010; Kana et al. 2015; Rosset 2008; Rosset and Rottman 2014; ter they experience them. As we predicted, participants were more Spunt, Meyer and Lieberman 2015), which has found that people likely to purchase the drawing in the mistake condition (M = 2.61, tend to assume that actions by agents (e.g., individuals, companies, SD = 1.73) compared to both the intention condition (M = 2.05, SD etc.) are done intentionally. We propose that a product made by mis- = 1.44; Fisher’s LSD: p = .014; d = 0.35) and the control condition take deviates from this assumption, and thus, is perceived as more (M = 2.13, SD = 1.53; Fisher’s LSD: p = .031; d = 0.29), which did improbable than when the same product is made with full intent. not differ from each other (Fisher’s LSD: p = .734). Willingness to In turn, this perceived improbability (for both products that are im- pay data revealed the same pattern, F(298) = 5.25, p = .006. Planned proved or worsened by a mistake) leads consumers to perceive the contrasts followed the same pattern, wherein participants were will- product as more unique and thus to prefer it. We test these predictions ing to pay more for the drawing in the mistake condition (M = 3.11, in a dataset of eBay auction sales, nine lab studies, and an additional SD = 3.28) compared to both the intention condition (M = 1.81, SD = study conducted in the field. We outline four of these studies below. 2.54; Fisher’s LSD: p = .002; d = 0.44), and the control condition (M = 2.28, SD = 2.79; Fisher’s LSD: p = .041; d = 0.27), which did not Conceptualization Of Mistake differ from each other (Fisher’s LSD: p = .257). The current research examines consumer preference for products made by mistake versus otherwise identical products made with Study 3 full intent. We define a mistake as “an error in action, calculation, Study 3 provided converging evidence for the proposed mecha- opinion, or judgment caused by poor reasoning, carelessness, in- nism via a moderation and mediation approach. Specifically, we pre- sufficient knowledge, etc.” (Random House Webster’s Unabridged dicted that if consumers are made aware that a particular product Dictionary 2001). Importantly, a mistake is defined by an outcome’s creator is likely to make a mistake, perceptions of the improbability cause (e.g., a careless deviation from intent) but not its valence (e.g., of that creator making a mistake will decrease and the preference whether the outcome is positive or negative). For instance, return- will attenuate. We tested this prediction in a 2 (Expertise: Novice ing to the example of the chef who mistakenly left the chocolate vs. Expert) × 2 (Creation Process: Mistake vs. Intention) between- in the oven for five extra minutes, this mistake could have a posi- subjects design. In the expert conditions, participants read that the tive outcome (the chocolate could be enhanced), negative outcome creator was a hip-hop producer who works in the recording indus- (the chocolate could be worsened), or neutral outcome (the chocolate try. In the novice conditions, participants read that the creator was could not be affected). Regardless of the outcome’s valence, the ex- a community college administrator taking a music production class, tra cooking time is a mistake because it is both a deviation from the and that the class assignments included doing recording sessions. We chef’s intent and the result of carelessness. Thus, in our conceptu- adapted this expertise manipulation from Karmarkar and Tormala alization of a mistake, consumers have to perceive the product as a (2010). Participants then learned that during a recent recording ses- result of an action that both deviates from the creator’s intent and is sion, the sound of the creator’s breath was added to a song. This ad- careless in nature. dition was either intentional (in the intention condition) or made by mistake (in the mistake condition) and it enhanced the overall sound and flow of the song.

As predicted, the interaction between expertise and creation by the perception that a photograph with a mistake is a more im- process on choice was significant (b = 1.29, p = .029). In the expert probable occurrence than a photograph made with full intent, and conditions, participants preferred the song when it was made by mis- thus is conferred with more uniqueness and value. It is interesting to take (67.4%) compared to when it was made intentionally (39.5%; note that photographs containing arguably negative properties (i.e., b = 1.15, p = 012). In the novice conditions, there was no difference blurriness) received a premium relative to those containing no such in song preference when it was made by mistake (49.1%) versus negative property. Thus, along with Study 2, these results may pro- intentionally (52.4%; b = .13, p = .721). Further analysis revealed vide converging evidence that the preference for products made by a significant interaction between expertise and creation process on mistake emerges not only in cases in which the mistake enhances the perceived improbability, F(1, 198) = 3.93, p = .049. Consistent with product, but also when it detracts from it. participants’ song decisions, the song was perceived as more improbable when an expert made it by mistake (M = 4.37, SD = 1.77) General Discussion compared to when the song was made intentionally (M = 3.47, SD Individuals and companies often avoid advertising their mis- = 1.41; Fisher’s LSD: p = .013; d = .56). In contrast, when a novice takes (Edmondson 1996; Michael 1976; Stefaniak and Robertson made the song, there were no differences between the mistake (M 2010; Uribe et al. 2002). However, we find that consumers some- = 4.26, SD = 1.75) and intention (M = 4.29, SD = 1.55) conditions, times prefer products made by mistake to otherwise identical prod- F < 1. The analysis also revealed the predicted interaction between ucts made intentionally. We identify the conditions under which this expertise and creation process on uniqueness, F(1, 198) = 7.67, p = preference emerges, and find that this preference is driven by the .006. The song was perceived as more unique when an expert made perception that a product made by mistake is more improbable than it by mistake (M = 4.45, SD = 1.03) compared to when the song was a product made without mistake. This perceived improbability in- made intentionally (M = 3.96, SD = 1.16) (Fisher’s LSD: p = .047; creases product uniqueness perceptions and subsequent preference. d = .45). In contrast, when a novice made the song, the opposite In support of this process, we find that the preference for prod- was the case—when the novice made the song intentionally it was ucts made by mistake manifests both when the mistake enhances deemed as marginally more unique (M = 4.44, SD = 1.22) than when (Study 1 & Study 3) and detracts from (Study 2 & Study 4) the prod- it was made by mistake (M = 4.04, SD = 1.08; Fisher’s LSD: p = uct. Study 3 provided moderation support for this process: When the .055; d = .35). creator had a high likelihood of making a mistake (a novice), the To test our proposed process, we ran a serial mediation model at preference for products made by mistake was attenuated. Study 3 each level of the expertise condition, with improbability and unique- also found full process evidence, wherein perceptions of the improb- ness as sequential mediators. As predicted, the model revealed that ability of the product’s creation enhanced perceptions of product when an expert made a mistake (vs. made the song intentionally), uniqueness and preference. Finally, Study 4 documented the exter- the song was perceived to be more improbable, which increased per- nal validity of the current results—consumers pay more for products ceived uniqueness, resulting in greater song choice (95% CI for the made by mistake in a real market setting. indirect effect: [-.7446, -.0465]). Importantly, we tested the reverse model with uniqueness preceding perceptions of improbability, and Theoretical Implications we found no mediating path from uniqueness and improbability lead- The current research is the first to reveal that consumers prefer ing to song choice (95% CI for the indirect effect: [-.2560, .0562]). products made by mistake, and that this preference emerges because Lastly, as predicted, for novices, there was no significant indirect mistakes are perceived to be improbable and the resulting product effect from improbability to uniqueness leading to song choice (95% more unique. Beyond documenting a new phenomenon, our find- CI for the indirect effect: [-.1157, .0264]). ings illuminate a novel mechanism through which information about the creation process influences consumer preference. We find that Study 4 consumers differentially prefer identical products depending on the Study 4 examined the preference for products made by mistake improbability of their creation. Further, we find that a typically nega- in a real market setting. We examined eBay auction sales of original tive occurrence (a mistake) actually increases product preference. photographs—some of which were made by mistake and some of This finding contributes to a growing body of work showcasing the which were not. The mistakes originated from a variety of causes, in- positive effects of negative information (Berger, Sorensen, and Ras- cluding double exposure, blurriness, and the photographer’s finger in mussen, 2010; Ein-Gar, Shiv and Tormala, 2012; Reich and Tormala, the exposure (i.e., finger bombing). Critically, all photographs were 2013; Reich and Wheeler, 2016). advertised as original, with no copies and reproductions of any kind Second, our work contributes to the underexplored role of in- available. In other words, all photographs (both those with mistakes tentions in consumer preference. By focusing on mistakes in par- and those without) were one-of-a-kind and hence equally rare. The ticular, we reveal a novel process by which intentionality plays a photographs were sold on eBay to the consumer who submitted the role in consumer preference. A product made by mistake deviates highest bid. We ran a regression on the price data with the mistake from the intentionally bias (Begue et al. 2010; Rosset 2008; Rosset dummy code as a predictor variable, and the sale date and photo- and Rottman 2014; Spunt et al. 2015), and consumers thus perceive graph size as control variables. The regression revealed a significant the process of making a product by mistake as more improbable and effect of photograph size on sale price, such that larger photographs hence more unique than the same product made with full intent. Fu- sold for more money (b = .529, p < .001). There was no effect of ture work could examine other potential consequences of the inten- sale date on sale price (b = -.008, p = .292). Most relevant to our tionality bias in the context of how consumers assess the actions of focal theorizing, the regression also revealed that photographs with individuals and companies. mistakes (vs. without mistakes) sold for more money (b = 6.08, p Finally, our findings also contribute to work on the role of < .001). In other words, photographs with a mistake sold for a pre- uniqueness in consumer preference. Prior work has focused on in- mium controlling for both photograph size and sale date. dividual differences in consumers’ need for uniqueness (Simonson Thus, we observe in an actual market context a premium for and Nowlis 2000; Tian et al. 2001) or on how uniqueness is distinct products with a mistake over products without a mistake. We posit, from other dimensions of value (Bhattacharjee and Mogilner 2014; in line with our experimental evidence, that this preference is driven Advances in Consumer Research (Volume 45) / 837

Keinan and Kivetz 2011). The present work illuminates a concep- Keinan, Anat, and Ran Kivetz (2011), “Productivity Orientation tual antecedent of perceptions of product uniqueness. Specifically, and the Consumption of Collectable Experiences.” Journal of we find that product uniqueness stems from the improbability of the Consumer Research, 37(June), 935-50. product’s creation. Our work suggests that these process features do Knobe, Joshua (2003), “Intentional Action and Side Effects in not necessarily have to positive, but rather unlikely. We encourage Ordinary Language.” Analysis, 63(279), 190-4. future research to examine other antecedents that might enhance Kruger, Justin, Derrick Wirtz, Leaf Van Boven, and T. William product uniqueness. Altermatt (2004), “The Effort Heuristic.”Journal of Experimental Social Psychology, 40, 91-8. REFERENCES Levin, Irwin P., and Gary J. Gaeth (1988), “How Consumers Are Bègue, Laurent, Brad J. Bushman, Peter R. Giancola, Baptiste Affected by the Framing of Attribute Information Before Subra, and Evelyn Rosset (2010), “There is No Such Thing as and After Consuming the Product.” Journal of Consumer an Accident: Especially When People are Drunk.” Personality Research, 15(3), 374-78.‏ and Social Psychology Bulletin, 36(10), 1301-04. Malle, Bertram F., and Joshua Knobe (1997), “The Folk Concept Berger, Jonah, Alan T. Sorensen, and Scott J. Rasmussen (2010), of Intentionality.” Journal of Experimental Social Psychology, “Positive Effects of Negative Publicity: When Negative 33(2), 101-21. Reviews Increase Sales.” Marketing Science, 29(5), 815-27.‏ Michael, Donald N. (1976), On Learning to Plan and Planning to Bhattacharjee, Amit, Jonathan Z. Berman, and Americus Reed Learn, San Francisco: Jossey-Bass. (2013), “Tip of the Hat, Wag of the Finger: How Moral Morales, Andrea C. (2005), “Giving Firms an “E” for Effort: Decoupling Enables Consumers to Admire and Admonish.” Consumer responses to High-Effort Firms.”Journal of Journal of Consumer Research, 39(6),1167-84. Consumer Research, 31(March), 806-12. Bhattacharjee, Amit, and Cassie Mogilner (2014), “Happiness Newman, George E., Margarita Gorlin, and Ravi Dhar (2014), from Ordinary and Extraordinary Experiences.” Journal of “When Going Green Backfires: How Firm Intentions Shape Consumer Research, 41(January), 1-17. the Evaluation of Socially Beneficial Product Enhancements.” Chesney, Thomas, and Daniel KS Su (2010), “The Impact of Journal of Consumer Research, 41(3), 823-39. Anonymity on Weblog Credibility.” International Journal of Nisbett, Richard E., and Timothy D. Wilson (1977), “Telling More Human-Computer Studies, 68(October), 710-18. than We Can Know: Verbal Reports on Mental Processes.” Edmondson, Amy C. (1996), “Learning from Mistakes is Easier Psychological Review, 84(3), 231-59. Said than Done: Group and Organizational Influences on the Palmer, Mark, Geoff Simmons, and Ronan de Kervenoael (2010), Detection and Correction of Human Error.” The Journal of “Brilliant Mistake! Essays on Incidents of Management Applied Behavioral Science, 32(January), 5-28. Mistakes and Mea Culpa.” International Journal of Retail & Ein-Gar, Danit, Baba Shiv, and Zakary L. Tormala (2012), “When Distribution Management, 38(March), 234-57. Blemishing Leads to Blossoming: The Positive Effect of Pizarro, David, Eric Uhlmann, and Peter Salovey (2003), Negative Information.” Journal of Consumer Research, 38(5), “Asymmetry in Judgments of Moral Blame and Praise the 846-59. Role of Perceived Metadesires.” Psychological Science, 14(3), Fuchs, Christoph, Martin Schreier, and Stijn MJ van Osselaer 267-72. (2015), “The Handmade Effect: What’s Love Got to Do with Random House Webster’s Unabridged Dictionary (2001), New It?” Journal of Marketing, 79(March), 98-110. York: Random House Reference. Gilbert, Daniel T., and Patrick S. Malone (1995), “The Reich, Taly, and Zakary L. Tormala (2013), “When Contradictions Correspondence Bias.” Psychological Bulletin, 117(1), 21-38. Foster Persuasion: An Attributional Perspective.” Journal of Haberstroh, Kristina, Ulrich R. Orth, Stefan Hoffmann, and Berit Experimental Social Psychology, 49(3), 426-39. Brunk (2015), “Consumer Response to Unethical Corporate Reich, Taly, and S. Christian Wheeler (2016),”The Good and Bad Behavior: A Re-Examination and Extension of the Moral of Ambivalence: Desiring Ambivalence When the Outcome Decoupling Model.” Journal of Business Ethics, 1-13. is Uncertain.” Journal of Personality and Social Psychology, Hassin, Ran R., Henk Aarts, and Melissa J. Ferguson (2005), 110(4), 493-508. “Automatic Goal Inferences.” Journal of Experimental Social Roberts, Royston M. (1989), Serendipity: Accidental Discoveries in Psychology, 41(2), 129-40. Science, New York: John Wiley & Sons. Heider, Fritz (1958), The Psychology of Interpersonal Relations, Roe, Brian, Alan S. Levy, and Brenda M. Derby (1999), “The New York: Wiley. Impact of Health Claims on Consumer Search and Product Hong, Sung-Tai, and Robert S. Wyer (1989), “Effects of Country- Evaluation Outcomes: Results from FDA Experimental Data.” of-Origin and Product-Attribute Information on Product Journal of Public Policy and Marketing, 18(1), 89-105. Evaluation: An Information Processing Perspective.” Journal Rosset, Evelyn (2008), “It’s No Accident: Our Bias for Intentional of Consumer Research, 16(2), 175-87. Explanations.” Cognition, 108(3), 771-80. Inesi, M. Ena, Simona Botti, David Dubois, Derek D. Rucker, and Rosset, Evelyn, and Joshua Rottman (2014), “The Big ‘Whoops!’ Adam D. Galinsky (2011), “Power and Choice: Their Dynamic in the Study of Intentional Behavior: An Appeal for a New Interplay in Quenching the Thirst for Personal Control.” Framework in Understanding Human Actions.” Journal of Psychological Science, 22(August), 1042-48. Cognition and Culture, 14(1-2), 27-39. Kana, Rajesh K., Carla J. Ammons, Constance F. Doss, Megan E. Rucker, Dereck D., and Adam D. Galinsky (2008), “Desire to Waite, Bhumika Kana, Abbey J. Herringshaw, and Lawrence Acquire: Powerlessness and Compensatory Consumption.” Ver Hoef (2015), “Language and Motor Cortex Response to Journal of Consumer Research, 35(August), 257-67. Comprehending Accidental and Intentional Action Sentences.” Neuropsychologia, 77(October), 158-64. 838 / Made by Mistake: When Mistakes Increase Product Preference

Ruvio, Ayalla, Aviv Shoham, and Maja Makovec Brencic (2008), Spunt, Robert P., Meghan L. Meyer, and Matthew D. Lieberman “Consumers’ Need for Uniqueness: Short-Form Scale (2015), “The Default mode of human brain function primes Development and Cross-Cultural Validation.” International the intentional stance.” Journal of Cognitive Neuroscience, Marketing Review, 25(1), 33-53. 27(6), 1116-24. Shiv, Baba, Ziv Carmon, and Dan Ariely (2005), “Placebo Effects Stefaniak, Chad, and Jesse C. Robertson (2010), “When Auditors of Marketing Actions: Consumers May Get What They Pay Err: How Mistake Significance and Superiors’ Historical For.” Journal of Marketing Research, 42(4), 383-93.‏ Reactions Influence Auditors’ Likelihood to Admit a Mistake.” Simonson, Itamar, and Stephen M. Nowlis (2000) “The Role of International Journal of Auditing, 14(March), 41-55. Explanations and need for uniqueness in consumer decision Tian, Kelly Tepper, William O. Bearden, and Gary L. Hunter making: Unconventional choices based on reasons.” Journal (2001), “Consumers’ Need for Uniqueness: Scale of Consumer Research, 27(1 (2000): 49-68. Development and Validation.” Journal of Consumer Research, Smith, Rosanna K., and George E. Newman (2014), “When 28(1), 50-66. multiple creators are worse than one: The bias toward single Uribe, Claudia L., Sharon B. Schweikhart, Dev S. Pathak, Gail authors in the evaluation of art.” Psychology of Aesthetics, B. Marsh, and R. Reed Fraley (2002), “Perceived Barriers Creativity, and the Arts, 8(March), 303-10. to Medical-Error Reporting: An Exploratory Investigation.” Smith, Rosanna K., George E. Newman, and Ravi Dhar (2016), Journal of Healthcare Management, 47(April), 263-79. “Closer to the Creator: Temporal Contagion Explains the Preference for Earlier Serial Numbers.” Journal of Consumer Research, 42(5), 653-68. Active Consumption: How the Architecture of the Experience Activates Consumer Engagement and Enjoyment Taly Reich, Yale University, USA Rosanna K. Smith, University of Georgia, USA Ernest Baskin, Saint Joseph’s University, USA

EXTENDED ABSTRACT Empirical Evidence Prior work has established the benefits of experiential over Study 1 material consumption on consumer wellbeing (Nicolao, Irwin and Study 1 examined the basic effect by varying whether architec- Goodman 2009; Gilovich, Kumar and Jampol 2014), however, little tural, neutral, or no information was provided prior to the same ex- attention has been devoted to the factors that enhance the very same perience. We predicted that the same experience would be enhanced experience. For instance, two individuals may attend the same art by the architectural information relative to the other two conditions. show or listen to the same song, but one might enjoy these experi- Specifically, in all three conditions, participants read that they would ences immensely, while the other feels disengaged. Holding constant listen to the song “Don’t Forget Me” by the Red Hot Chili Peppers. the experience, what factors influence the extent to which consumers In the architecture condition, before listening to the song, partici- more or less enjoy it? pants read about the structure of the song (i.e., the chords). To control In this article, we focus on the role of providing the architecture for cognitive load, participants in the neutral information condition, of the experience on subsequent consumer engagement and enjoy- prior to listening to the song, read a paragraph of equivalent length ment. Past research suggests two competing intuitions about the in- on a neutral topic. Finally, in the control condition, participants pro- fluence of prior information more generally on consumer enjoyment ceeded to listen to the song with no further information provided. of an experience. On the one hand, revealing information could at- Participants in all three conditions then listened to the song and tenuate the enjoyment by both disrupting flow and reducing potential reported how much they had enjoyed listening to it. As predicted, for surprise (Csikszentmihalyi, 1975). On the other hand, other work planned contrasts revealed that participants enjoyed the song more has suggested that prior knowledge can increase enjoyment by en- when they were given information about the architecture of the song hancing consumer involvement (Bloom 2010; Humphrey, Thomas, (M = 5.56, SD = 1.41) compared to when they were given unrelated and Joshua Gutwill 2005). information (M = 4.80, SD = 1.85) and when they were given no ad- To reconcile these competing hypotheses, we propose that cer- ditional information (M = 5.12, SD = 1.75), t(300) = -2.86, p = .004, tain kinds of knowledge (as opposed to knowledge more generally) d = 0.38. Further, the neutral information and control condition not can enhance experiences. Specifically, we propose that information differ from each other, t(300) = 1.38, p = .17. Of note, the effect of about the architecture of an experience uniquely enhances enjoy- condition on enjoyment did not interact with prior familiarity with ment relative to other forms of knowledge. Architectural informa- the song (p = .23). tion is defined as knowledge about the structure, arrangement, and/or framework of the experience (adapted from Fischer, Winter and Aier Study 2 2010; Winter, Buckl, Matthes, and Schweda 2010). For instance, in a In Study 2, we were interested in examining whether people song, the architectural information would be the arrangement of the intuit that information about the architecture of the experience will chords and verses, but not the actual lyrics. In this way, architectural enhance their consumption experience. To test this, we asked par- information is not about the content per se of the experience, but spe- ticipants to predict how information about the architecture of a song cifically about the organization or outline of it. Further, architectural would influence how others would enjoy the song. Specifically, par- information is embedded within the experience itself. ticipants were asked to indicate whether people who were provided We propose that this combination of features leads architec- with information about the structure of a song would enjoy it more tural information to enhance experience in contrast to other kinds of relative to those who were given no information about the structure. knowledge. Because architectural information is embedded within Consistent with our hypothesis, most participants (68%) predicted the experience itself, consumers are cued to seek it as the experience that the song would be enjoyed more in the no information condition unfolds, at the same time, because only architectural information, compared to the architecture information condition (χ2(1) = 29.42, p and not content, has been revealed, the potential for surprise is main- < .001). In addition, this pattern of results did not vary as a function tained. In some sense, architectural information is in a “sweet spot” of familiarity with the song (χ2(1) = .86, p = .355). Thus, we find where it is just enough information to induce engagement but not that people do not intuit that architectural information will enhance enough to induce boredom. enjoyment. In the following studies, we test the basic hypothesis that ar- Study 3 chitectural information enhances an experience relative to the same Study 3 had two aims. The first was to test our proposed pro- experience without prior architectural information. We then test peo- cess. Specifically, we proposed that information about architecture ple’s intuitions about the influence of architectural information on enhances enjoyment because it prompts consumers to actively seek their experience. The last two studies employ a mediation and a mod- it within the experience itself. This “structure-seeking” increases ac- eration approach respectively in order to test our proposed process. tive participation in the experience and subsequently enhances en- Specifically, we test whether exposure to architectural information joyment. The second aim was to test whether this active participation increases active participation and subsequent enjoyment of the expe- is uniquely related to architectural information as opposed to general rience. Finally, we contrast the effect of architectural information to knowledge about the experience. To do so, we also manipulated the other forms of knowledge about the experience. type of knowledge about the experience (structure vs. general knowledge).

Participants were randomly assigned to one of three conditions: entitled “Summer Strut.” To directly measure our proposed process, architecture, general song information and control. As in Study 1, we again asked participants to complete the structure-seeking item as in all conditions participants read that they would listen to the song in Study 3. Further, participants also completed the same two active “Don’t Forget Me” by the Red Hot Chili Peppers. In the architec- engagement items (α = .80). ture condition, participants were provided with the same structure As predicted, planned contrasts revealed participants enjoyed information as in Study 1. In the general song information condition, the song more in the architecture before song condition (M = 4.13, participants read information about the band. In the control condi- SD = 1.49) compared to both the architecture after song (M = 3.69, tion participants listened to the song with no further information pro- SD = 1.47) and the control (M = 3.58, SD = 1.28) conditions, t(242) vided. Next, participants listened to the song and reported how much = -2.60, p = .01, d = .35, which did not differ from each other,t (242) they enjoyed it using the same measure as in Study 1. To directly test =.51, p = .61. Replicating the results of Study 3, participants in the our proposed process, participants then completed three randomized architecture before song condition reported seeking structure more items. To capture structure-seeking, participants were asked to rate (M = 3.94, SD = 1.91) compared to both the architecture after song the extent to which they looked for structure when listening to the (M = 3.28, SD = 1.84) and the control (M = 3.45, SD = 1.76) condi- song (on a scale from 1:Not at all; 7:Very much). To assess active en- tions, t(242) = -2.33, p = .02, d = .31, which did not differ from each gagement, participants responded to the following two items: “When other, t(242) = -.56, p = .57. Similarly, active engagement mirrored you were listening to the song, to what extent did you feel you were the structure-seeking pattern, with participants in the architecture an active participant?” (1: Passive - 7: Active); “When you were lis- before song condition indicating higher active engagement with the tening to the song, to what extent did you feel you were an active song (M = 4.16, SD = 1.53) compared to both the architecture after listener?” (1: Not at all - 7: Very much). These two items were then song (M = 3.80, SD = 1.54) and the Control (M = 3.67, SD = 1.34) combined into an active engagement composite (α = .69). conditions, t(242) = -2.13, p = .03, d = .29, which did not differ from Consistent with our theorizing, planned contrasts revealed that each other, t(242) =.54, p = .59. participants enjoyed the song more in the architecture condition (M We conducted a serial mediation model with bootstrapping = 5.67, SD = 1.43) compared to both the general information (M = (Hayes 2013), to test whether structure seeking and subsequent ac- 4.82, SD = 1.64) and the control (M = 4.98, SD = 1.63) conditions, tive engagement mediate the relationship between condition and t(167) = -3.04, p = .003, d = .51, which did not differ from each other, enjoyment. The model revealed that indeed, participants in the ar- t(167) = .54, p = .59. Further, in line with our proposed mechanism, chitecture before song condition sought structure more, which led participants in the architecture condition reported seeking structure them to be more actively engaged with the song, resulting in greater more (M = 5.24, SD = 1.54) compared to both the general informa- enjoyment (95% CI for the indirect effect: [.0236, .2895]). tion (M = 4.18, SD = 1.69) and the control (M = 4.24, SD = 1.69) conditions, t(167) = -3.89, p < .001, d = .64, which did not differ from General Discussion each other, t(167) = .18, p = .86. Further, active engagement paral- Across four studies, we find that the very same experience is leled the structure-seeking pattern, with participants in the architec- enjoyed differently depending on what information people are ex- ture condition indicating higher active engagement with the song (M posed to prior to consumption. Specifically, we find that architectural = 5.71, SD = 1.13) compared to both the general information (M = information (as opposed to general knowledge or no knowledge) en- 5.03, SD = 1.39) and the control (M = 4.72, SD = 1.66) conditions, hances enjoyment. Because this information is embedded within the t(167) = -3.70, p < .001, d = .62, which did not differ from each other, experience itself, consumers are prompted to seek the structure of the t(167) = -1.17, p = .25. experience as it unfolds. This, in turn, increases active participation Finally, we conducted a serial mediation model with bootstrap- in the experience and subsequent enjoyment. ping (Hayes 2013), to test whether structure seeking and subsequent Our work has several theoretical and practical implications. active engagement mediate the relationship between condition and While prior research has focused on the contrast between experi- enjoyment. We predicted that participants who were given the ar- ential versus material consumption or on comparing different kinds chitecture before the song would enjoy the song more because they of experiences (Bhattacharjee and Mogilner 2013; Nicolao, Irwin would be more likely to seek the structure, subsequently increasing and Goodman 2009; Gilovich, Kumar and Jampol 2014), little atten- their engagement. The model revealed that indeed, participants in the tion has been devoted to what factors differentially enhancethe very architecture condition sought structure more, which led them to be same experience. While experiences maybe more enjoyable relative more actively engaged with the song, resulting in greater enjoyment to material goods, consumers do not always find experiences engag- (95% CI for the indirect effect: [.0478, .2730]). Thus we confirm ing. Our work explores a novel antecedent to enhancing experiential our process and show that active participation is uniquely related to consumption—prior information about the architecture of the expe- structure seeking as opposed to exposure to prior knowledge more rience. Future work could further unpack other antecedents of active generally. engagement in experiences. Companies and marketers can use this work to be more cog- Study 4 nizant of what kinds of information to reveal to consumers prior to In Study 4, we tested our process using a moderation approach consumption. Our work suggests that consumers do not always have by having participants learn about architecture either before or after the most accurate intuitions about what information or knowledge the experience. Specifically, participants were randomly assigned to will enhance their experiences. Understanding what facilitates con- one of three conditions: architecture before song, architecture after sumers to move from passive to more active consumption is a fruitful song, and control. Further, for this study, we created an original song area for researchers and practitioners alike. in order to ensure that participants would have no prior familiarity or association with the song. The new song was created using a soft- REFERENCES ware called JamStudio (http://www.jamstudio.com/Studio/indexStu- Bhattacharjee, Amit, and Cassie Mogilner (2014), “Happiness dio.htm), which is an online interface that allows the user to select from Ordinary and Extraordinary Experiences.” Journal of chords and vary the instruments to create a song. In all three condi- Consumer Research, 41(June), 1-17. tions, participants were told that they were going to listen to a song Advances in Consumer Research (Volume 45) / 841

Bloom, Paul (2010), How Pleasure Works: The New Science of Why Humphrey, Thomas, and Joshua Gutwill (2005), Fostering Active We Like What We Like, New York: W.W. Norton & Company. Prolonged Engagement: The Art of Creating APE Exhibits, Csikszentmihalyi, Mihaly (1975), “Play and Intrinsic Rewards.” San Francisco, CA: Exploratorium. Journal of Humanistic Psychology, 15(3), 41-63. Nicolao, Leonardo, Julie R. Irwin, and Joseph K. Goodman Fischer, Christian, Robert Winter, and Stephan Aier (2010), “What (2009), “Happiness for Sale: do Experiential Purchases Make is an Enterprise Architecture Principle?” In Computer and Consumers Happier than Material Purchases?” Journal of Information Science 2010, 317, 193-205. Consumer Research, 36(2), 188-98. Gilovich, Thomas, Amit Kumar, and Lily Jampol (2015), “A Winter, Katharina, Sabine Buckl, Florian Matthes, and Christian Wonderful Life: Experiential Consumption and the Pursuit of M. Schweda (2010), “Investigating the State-of-the- Happiness.” Journal of Consumer Psychology, 25(1), 152-65. Art in Enterprise Architecture Management Methods in Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, Literature and Practice.” in 5th Mediterranean Conference and Conditional Process Analysis: A Regression-based on Information Systems, Tel Aviv, Israel: MCIS 2010 Approach. New York: Guilford Press. Proceedings, 90. Praise for Blame: Consumer Inferences following Cause Marketing Brandon Reich, University of Oregon, USA Troy Campbell, University of Oregon, USA Robert Madrigal, California State University at Chico, USA

EXTENDED ABSTRACT scores represented greater fit. Regressing praise on blame fit showed Companies often engage in cause marketing (CM), defined the predicted negative relationship (b=-.69, SE=.10, p<.001), which here as any corporate communication of a company’s prosocial ac- remained significant (p<.001) when controlling for the covariates. tivities. A substantial literature is dedicated to exploring how CM These results suggest that inferred blame may determine praise for produces positive consumer outcomes (see Peloza and Shang 2011), the company depending on how well it fits consumers’ own blame with recent focus on consumers’ inferences of the company’s mo- judgments. tives (Reich and Armstrong Soule 2016; Sen, Du, and Bhattacharya Study 1B (N=92 MTurk workers; MAge=36.70, SDAge=12.60; 2016). We diverge from extant research in this regard, examining 47.8% female) was identical to study 1A, except that blame mea- consumer inferences around blame. We propose and show that CM sures were dichotomized (0=Congress, 1=Violent Individuals). We may be implicitly communicating blame for several blame targets tested an inferred blame × consumer blame interaction on praise, and around the issue. Crucially, the extent to which consumers perceive results further supported the blame fit hypothesis (b=2.22, SE=1.23, this blame may benefit or harm to the company depending on how p=.074; see the figure, panel B). We also created a new blame fit well it fits with consumers’ own blame for the same issue (i.e., con- variable, coded as 1 when participants selected the same target for sumer-company “blame fit”). both blame questions, and 0 when they did not. As expected, praise In establishing these effects, we contribute to several relevant was greater among those who perceived blame fit (n=53, M=6.47, literatures. First, we reconceptualize and apply blame in novel ways. SD=2.73) than among those who did not (n=39, M=4.77, SD=2.78, Specifically, we show that blame may be communicated through t(90)=2.935, p=.004). This relationship held (p=.014) while control- CM and inferred by consumers, irrespective of a company’s intent ling for the covariates. to blame. Second, we introduce a novel construct, blame fit, and es- Study 1C (N=185 MTurk workers; MAge=33.01, SDAge=8.61; tablish its positive relationship with praise for the company. To our 42.2% female) used a manipulation of explicit blame, testing wheth- knowledge, this is the first demonstration of blame as a praisewor- er blame fit effects would manifest similarly as in cases of inferred thy action. Last, we extend the scope of theoretical models of blame blame. The procedure was similar to studies 1A and 1B, except the ad (e.g., Alicke 2000; Folkes, 1984; Klein and Dawar 2004; Malle, included an additional block of text reading either “Unregulated sale Guglielmo, and Monroe 2014), showing that an observer (e.g., con- of guns harms us all” (blame Congress condition) or “Gang violence sumer) may infer blame from another blame agent (e.g., company). harms us all” (blame violent individuals condition). Perceived blame In a pilot study (N=75 undergraduate business students; (as manipulation checks) and consumer blame for these two targets were measured as in study 1A. Supporting the effectiveness of the MAge=21.24, SDAge=2.34; 33.3% female), participants viewed an ad for an environmentally-friendly Toyota model, and reported whether manipulation, those in the blame Congress (individuals) condition they thought Toyota was blaming someone or something for environ- perceived greater company blame for Congress (violent individuals; mental problems (0=No; 1=Yes). More than half of the sample (n=45, ps<.001). Moreover, a condition × consumer blame interaction on 60.0%, t(74)=1.756, p=.083) responded affirmatively. The company praise was observed when treating consumer blame for Congress did not explicitly blame anything, but the ad led a substantial propor- as the moderator (b=.21, SE=.11, p=.057; see the figure, panel C), tion of consumers to infer blame. but not when consumer blame for individuals was the moderator (p=.844). This study demonstrated that inferred blame may function Study 1A (N=200 MTurk workers; MAge=38.99, SDAge=13.06; 58.0% female) tested the blame fit hypothesis, i.e., that greater fit be- as though it were explicit blame from the company, and that blame fit tween inferred blame and consumers’ own blame increases praise for affects praise similarly in both cases. the company. Participants viewed a fictional Ben & Jerry’s ad depict- A pilot study established the existence of inferred blame in re- ing their cow mascot holding a sign with a twisted gun and reading sponse to CM, and three follow-up studies showed that the relation- “Stop the Violence.” On nine-point semantic differential scales, they ship between blame fit and praise appears robust. Pragmatically, this rated (1) perceptions that Ben & Jerry’s was blaming four separate implies that CM may be communicating more than intended. Rather targets (violent media, the NRA, Congress, and violent individuals) than leaving blame perceptions to consumers’ inferences, companies for gun violence, (2) their own blame for these four targets, and (3) may benefit from explicitly blaming one target or another, depend- their praise for the company. Measured covariates were pre-existing ing on the target audience. This research also expands blame theory attitudes toward Ben & Jerry’s, urge to help, issue seriousness, and to include novel constructs (“inferred blame” and “blame fit”), out- political orientation. comes (blame as a praiseworthy action), and perspectives (beliefs We first tested interactions between inferred company blame about another agent’s blame). Additional research may build on the and corresponding consumer blame for the four blame targets, with present findings by examining boundary conditions to the observed praise as the dependent variable. Results supported the blame fit effects. For instance, the presence or absence of a possible causal hypothesis via significant interactions on praise when treating the agent (Malle et al. 2014) may determine whether any blame is in- NRA (b=.07, SE=.02, p<.001), Congress (b=.09, SE=.02, p<.001), ferred. and violent individuals (b=.07, SE=.04, p=.060) as the blame target (see the figure, panel A), but not when violent media was the REFERENCES blame target (p=.881). As an alternative test, we created an index Alicke, Mark D. (2000), “Culpable Control and the Psychology of of overall blame fit by calculating the absolute value of the differ- Blame,” Psychological Bulletin, 126 (4), 556-574. ence between inferred blame and consumer blame for each target, and averaging these four difference scores (α=.653) such that lower

Folkes, Valerie S. (1984), “Consumer Reactions to Product Failure: Reich, Brandon J., and Catherine A. Armstrong Soule (2016), An Attributional Approach,” Journal of Consumer Research, “Green Demarketing in Advertisements: Comparing “Buy 10 (4), 398-409. Green” and “Buy Less” Appeals in Product and Institutional Klein, Jill G., and Niraj Dawar (2004), “Corporate Social Advertising Contexts,” Journal of Advertising, 45 (4), 441- Responsibility and Consumers’ Attributions and Brand 458. Evaluations in a Product–Harm Crisis,” International Journal Sen, Sankar, Shuili Du, and Chitra Bhanu Bhattacharya (2016), of research in Marketing, 21 (3), 203-217. “Corporate Social Responsibility: A Consumer Psychology Malle, Bertram F., Steve Guglielmo, and Andrew E. Monroe Perspective,” Current Opinion in Psychology, 10, 70-75. (2014), “A Theory of Blame,” Psychological Inquiry, 25 (2), 147-186. Peloza, John, and Jingzhi Shang (2011), “How can Corporate Social Responsibility Activities Create Value for Stakeholders? A Systematic Review,” Journal of the Academy of Marketing Science, 39 (1), 117-135. Using a Neural Network Model to Assess Advertising Effectiveness: A Validation of the Strategy Assessment (Strata) Model Thomas Reynolds, (Emeritus) University of Texas, Dallas, USA

EXTENDED ABSTRACT el underlying neural network theory. That is, to validate the neural In this paper we draw a parallel between means-end decision network-based Strata model, we empirically assess if the neural con- theory and neural network analysis and apply this common per- necting linkages caused by a given advertisement are related to an spectives to empirically validate an advertising strategy assessment increase in the likelihood of purchase. model with respect to predicting purchase intent. The results of the meta-assessment of 240 television ads offer strong support for the Background neural network-based model. Neural Networks A neural network is a “connectionist” model of brain behav- Introduction ior often used to understand human cognition (West et al., 1997). Neuroscience is an emerging field of consumer research that has The interconnections, or linkages between neurons, are referred to garnered much interest among advertising researchers in the hope as “synapses.” Neural network models resemble the brain’s decision- that neuroscience tools can help them better understand why cus- making process where input neurons receive stimuli which are then tomers prefer some products over others (Nobel, 2012; Plassmann fed into a pattern matching process yielding a decision (Bhatt, 2012; et al., 2012). This ambitious goal relies on several neuro-, or brain Curry and Moutinho, 1993). The fundamental principle of a neural imaging, techniques deployed to reveal the hidden elements of the network is that when a neuron is activated, or fired, it can then cross consumer decision process to better understand how a product or a synapse gap to activate another ‘connected’ neuron. Put simply, an ad engages the pleasurable reward center in consumers’ brains neural networks models are computational models intended to repre- (Karmarkar 2011; Nobel 2012; Smidts et al., 2014). Advantages of sent biological neural networks in the brain and are used by research- neuromarketing approaches are that they are potentially faster and ers to solve certain kinds of problems (e.g., Briesch and Rajagopal, cheaper than traditional advertising research tools that ask customers 2010; Chowdury and Samuel, 2013). The basic logistical calculus directly for their thoughts, feelings, and decision-making strategies of a neural network is that a neuron receives inputs, processes those (Ariely and Berns, 2010). Although cost savings is one possible out- inputs, and generates an output (McCulloch and Pitts, 1943). In gen- come of this evolving area of research, most experts believe that the eral, neural networks do not follow a linear path as it is believed that potential contribution of neuroscience to advertising is its ability to information is processed collectively throughout the entire network guide theory generation that can be used to shape new models of con- of neurons, also called nodes. sumer decision making, and its ability to assess and/or supplement It is worth noting that with respect to decision making Hebb’s traditional models of consumer responses to advertising currently in (1949) synaptic plasticity postulate suggests that one’s neural net- practice (Ariely and Berns, 2010; Plassmann et al., 2012; Smidts et work is strengthened over time, becoming more efficient and effica- al., 2014; Yoon et al., 2012). cious as a result of repeated activation (i.e., stimuli exposure) and Although the importance of neuroimaging studies of advertis- personal experience. Thus, neural network analysis can be used to ing is expected to increase, to date neuroimaging’s relevance for understand consumer decision making with respect to a problem that practice has been limited (Ariely and Berns, 2010; Smidts et al., a product or service solves and subsequently offers the potential to 2014). Studies that have used neuroimaging to predict consumer assess advertising effectiveness. choice in response stimuli have not been found to be more predictive In the next section we introduce a well-known consumer de- than consumers’ own self-reports (Knutson et al., 2007; Plassmann cision-making model, means-end theory (Gutman, 1982) and draw et al., 2012). parallels to a neural network model of decision making. One area of neuroscience that has advanced our understanding of consumer decision making is neural network research which was Means-End Theory and Research Methods inspired by the neural architecture of the human brain (West et al., A commonly utilized and frequently-referenced framework of 1997). The strength of this approach is its ability to mimic the brain’s consumer decision making is means-end theory (Gutman, 1982) and function. From a cognitive perspective, neural network models are its associated research methodology, laddering (Reynolds and Gut- consistent with a spreading activation model of memory making man, 1988; Reynolds and Phillips, 2009). The laddering methodol- neural network models well-suited for representing judgment and ogy begins with a trained interviewer asking a series of questions to decision making that involves the processing of information (Bhatt, a consumer with the goal of abstracting to the higher-order mean- 2012; Chowdhury and Samuel, 2014; Payne et al., 1993; West et al., ings that drive the consumer’s decision making. The first questions 1997). Therefore, we posit that neural network research is a particu- of a laddering interviewer usually elicit a distinction between two larly apt approach to better understand advertising effectiveness (Bri- stimuli, or choice options, (e.g., Most preferred brand [Starbucks] esch and Rajagopal, 2010; Curry and Moutinho, 1993). vs. Second choice [Illy]), with regard to a stated preference (i.e., why In this paper we apply a neural network approach to assess a do you prefer Starbucks to Illy?) or actual consumption (i.e., why do theoretical model of consumer decision making used in advertising you drink more Starbucks than Illy?). Then the interviewer probes research, the Strata model (Reynolds, Gengler and Howard, 1995; the consumer’s answers with some version of the question, Why is Reynolds and Rochon, 1991). Specifically, we report on a meta-as- that important to you? The interviewer uses each answer as the ba- sessment of 240 television ads across a variety of product categories sis for the subsequent probe moving the consumer up the “ladder of and levels of finish (finished ads and animatics) that served as stimuli abstraction” from [1] Attribute distinction to [2] Functional Conse- for diverse samples of consumers. Study participants responded to quences to [3] Psycho-Social Consequences to [4] Personal Values. questions via a computer-driven, tailored interviewing system such The laddering methodology’s levels of abstraction are isomor- that the order of questioning reflected the spreading activation mod- phic to the neural network approach of obtaining the relevant connec-

Advances in Consumer Research 844 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 845 tions both between and across the decision-based nodes. The result a 0,1 or 2 score for each node and their linkage, meaning of a laddering interview is a complete means-end chain consisting the maximal linkage score possible for each connection is of four concepts (i.e., nodes) and three adjacent, direct connections 8. The likelihood of possible scores is used to convert the (i.e., linkages) [1-2; 2-3; 3-4] as well as indirect connections [1-3; multiplicative score to a 0-9 connection score. 1-4; 3-4]. The end result of a laddering study is a directed graph depicting the “network” of direct and indirect connections across nodes • The Affect ratings are scored in the same manner as the for a given sample of consumers. This graphical network is called a node ratings (0-100 scale), with a slightly modified ques- Consumer Decision Map (CDM) as it illustrates the consumers’ key tion format. decision nodes and their dominant associative connections. It is worth noting that means-end theory may be viewed as a Meta-Assessment Framework top-down approach to understanding consumer decision making, The Strata model permits the assessment of the validity of inter- while the laddering methodology is bottom-up whose goal is to iden- preting the linkage connectivity as a neural network, both in relation tify the end-state that defines the motivating dynamic of the deci- to the correspondence of an individual level and overall combining sion structure. That is, the lower levels develop their importance by all three linkages (Low=Product Bridge; Middle=Personal Rel- satisfying the respective, adjacent higher levels, while the laddering evance Bridge; Highest=Value Bridge), to the advertising effective- methodology is initiated by a distinction that is typically at the low- ness dependent measure of ‘Motivates Purchase’. est, attribute level. Thus, the goal of the laddering methodology is to Sample uncover a network of meanings, which also defines the association The 240 ads in the sample were from eight countries with network of connections, or linkages. Thus, the CDM may be viewed roughly half from the United States. There were 131 finished ads as a special case of neural network model: one that focuses on only and 109 animatics in the sample. Overall, the respective number of the aggregate, relevant decision-making elements assuming that the ads for each of the four general ‘product’ categories of ads were: levels of abstraction reflect the underlying decision-making process. n=131 beverages (including both alcoholic and non-alcoholic); 56 Advertising Effectiveness automobiles; 27 trade organizations; and, 26 ads across a diverse Reynolds and Rochon (1991) draw a meaningful distinction be- variety of consumer goods. The average number of the sample for tween traditional copy testing methods and advertising strategy as- each ad assessment was 42. The respondent sample composition was sessment which they operationalize as a neural network between the typically a combination of ‘most often’ brand users and ‘competitive means-end levels of the MECCAS framework (Reynolds and Crad- brand most often’ users, balancing for gender and age. For the trade dock, 1988). They suggest a model which is driven by a computer organization samples only gender and age were used. presentation of questions relating to what the ad caused them to think 2 about that are tailored to the respondent depending on their prior an- Findings: Correlations and R with Purchase Motivation swers that is comprised of: [1] affect-related questions (both Product Table 1 summarizes the zero order correlations with Purchase and Ad), [2] assessment of the strength of communication of ‘nodes’ Motivation for 240 ads with respect to the three linkage scores lev-   by level (the statements are developed from prior laddering research) els (L M H) representing the largest nodes-in-common ‘ladder’ and [3] the level of association between the adjacent-level nodes that connecting from Lowest (Message Element) to Highest (Personal were ‘clearly communicated’. During this tailored interview each ad Value) and the simple sum of the three levels, which represents a is viewed four times. Noteworthy is that the Strata methodology re- ‘neural-based’ strategy model s[L+M+H], compared to the tradi- quires that two ads are assessed by the respondent at the same time tional copy testing proxy of Ad Affect. The square of the correlation during the interview session to avoid the respondent over-analyzing representing the variance accounted for can be viewed to contrast a given ad (thereby avoiding the ‘expert effect’). The summary quan- the magnitude of the difference between the neural-based approach titative output provided for a given advertisement across a sample of to quantifying overall strategy assessment to traditional copy testing. respondents. The likelihood of the statistical difference between the correlations can be summarized by a z-score reflecting likelihood, z-[S-A], based • Average node (0-100) strength with a two-step rating [a] upon the ad sample size. if the node concept is ‘clearly communicated’ or not, and Table 1 . Correlations of Connections and Ad [b] if the node concept is determined to be ‘clearly’ com- Affect with Purchase Intent (Motivation) municated the respondent is asked if it is ‘perfectly com- #Ads Lowest Middle Highest S[L+M+H] Ad Affect z-[S-A] municated’ or (just) ‘clearly’ (Perfectly is scored as a 100, Overall 240 r .70 .68 .61 .78 .45 6.1 r*2 .49 .47 .37 .62 .21 Clearly is scored as 62 and Not Clearly is 0). (Note: for the predetermined key strategic elements if the concept The correlations by linkage level with Purchase Motivation are, was ‘clearly communicated’, the respondent is asked what of course, all statistically significant. And, the difference between the specifically in the ad was communicated that led to endors- Lowest level of .70 and the Highest of .61 is not statistically differing that concept statement. Analysis of this qualitative data ent. The key overall model correlational contrast of the s[L+M+H] of has used to understand what executional cues activate the .78 to Ad Affect of .45 reveals that the neural-based strategy model higher level meanings.) accounted for almost three times the variance as the traditional copy testing approach. The likelihood of that difference is represented by • Linkage strength is computed by asking for only those the z-score of 6.1. In sum, the neural framework underlying the deci- nodes that are ‘clearly communicated’ between adjacent sion process is validated. levels the degree of ‘associative meaning’ caused by the ad. A Venn diagram defines the graphical rating scale. Research Implications and Future Directions The computation of the summary index is based upon a Given that the Strata advertising assessment measures of neural probabilistic function derived from multiplicative model connection are validated in terms of quantifying their effect on deci- of Node i x Node ii x Linkage i-ii, which is computed from sion making, the general means-end approach grounded in under- 846 / Using a Neural Network Model to Assess Advertising Effectiveness: A Validation of the Strategy Assessment (Strata) Model standing the consumer bases of competitive preferences to develop Chowdhury, P. and M.S. Samuel. “Artificial neural networks: a advertising strategy is also validated. There are several implications tool for understanding green consumer behavior.” Marketing of these findings. Intelligence & Planning, 32, 5 (2014): 552-566 Curry, B. and L. Moutinho. “Neural Networks in Marketing: Formalizing the Advertising Creative Development Modeling Consumer Responses to Advertising Stimuli.” Challenge European Journal of Marketing, 27, 7 (1993): 5-20. Given that these analyses strongly suggests the key to effective Gengler, C. and T.J. Reynolds. “Consumer Understanding and advertising is to cause the associative connections between the four Advertising Strategy, Analysis, and Strategic Translation of nodes, asking the creative team to specify as precisely as possible Laddering Data.” Journal of Advertising Research, 35 (1995): what exactly in a proposed execution will result in the linkages being 19-33. made provides a process to focus meaningful discussion and ongo- Gutman, J. (1982). “A Means-end Chain Model based on Consumer ing learning. (Gengler and Reynolds, 1995). Categorization Processes.” Journal of Marketing, 46 (1982): Advances in Laddering Methods 60-72. The central role of preference-based laddering to develop stra- Hebb, D. The Organization of Behavior. New York: Wiley & Sons, tegic positioning options, as well as the development of product- 1949. specific statements to be used in Strata for decades has involved Karmakar, U.A. Note on Neuromarketing. Harvard Business one-on-one in-depth interviews. This significant time and cost limi- School, 2011. tation of this standard approach to laddering has been addressed by Knutson, B., S. Rick, G.E. Wimmer, D. Prelec, and G. Loewenstein. an internet-based, one-on-one interviewing system termed Stream, “Neural predictors of purchases.” Neuron, 53, 1 (2007): 147- which has been shown to produce significantly higher quality data in 156. a significantly more efficient manner (Reynolds and Phillips, 2009). McCulloch, W.S. and W.H. Pitts. “A Logical Calculus of the Ideas The Stream software also contains an on-line coding function which Immanent in Nervous Activity.” Bulletin of Mathematical greatly facilitates this tedious process, along with a decision seg- Biophysics, 5 (1943): 115-133. mentation methodology (Reynolds, 2006) which permits straight Nobel, C. What Neuroscience Tells Us About Consumer Desire. forward contrasts with traditional types of marketing research clas- HBS Working Knowledge, 2012 sification data and facilitates the strategy development process (Phil- Payne, J.W., J.R. Bettman, and E.J. Johnson. The Adaptive Decision lips, Reynolds, and Reynolds, 2010). Maker. New York: Cambridge University Press, 1993 The likely next research advance will be the use of artificial Phillips, J.M., T.J. Reynolds, and K. Reynolds. “Decision-Based intelligence (AI) software to conduct decision-based laddering in- Voter Segmentation: An Application for Campaign Message terviews. The author is aware of one such system which has shown Development”, European Journal of Marketing, 44 (2010): promise. Interestingly, because the software teaches the respondent 310 – 330. how means-end-based decision making functions, including how Plassmann, H., T.Z. Ramsoy, and M. Milosavljevic. “Branding the choice trade-offs are framed (Reynolds, 2005), the AI approach to Brain: A Critical Review and Outlook.” Journal of Consumer laddering results in very high respondent involvement with the ques- Psychology, 22 (2012): 18-36. tioning process. Reynolds, T.J. “LifeGoals: The Development of a Decision-Making Curriculum for Education.” Journal of Public Policy and Summary Marketing, 24 (2005): 75-81. To assess the neural theoretic underlying the identification Reynolds, T.J. and A. Craddock. “The Application of the MECCAS of the decision network activated by advertising the Strata model Model to the Development and Assessment of Advertising (Reynolds and Rochon, 1991) was used. A meta-analysis of 240 Strategy: A Case Study.” Journal of Advertising Research 28 ads was analyzed yielding data measuring the strength of cogni- (1988): 43-54. tive association linkages (neuroscience) between levels abstraction Reynolds, T.J. and C. Gengler. “The Strategic Assessment of (Means-end theory) caused by advertising stimuli. This aggregate Advertising: The Animatic Versus Finished Issue.” Journal of data is seen to be highly related on a correlational basis to advertising Advertising Research, 31 (1991): 61-71. effectiveness, operationalized as the level of Purchase Motivation for Reynolds, T.J., Gengler, C. and D..Howard. “A Means-end Analysis the advertised product. of Brand Persuasion in Advertising.” International Journal of The critical finding of this meta-analysis is that a decision- Research in Marketing, 12 (1995): 257-266. based methodological platform for understanding advertising effec- Reynolds, T.J. and J. Gutman. “Laddering Theory, Method, tiveness defined by directional changes in Purchase Motivation has Analysis and Interpretation.” Journal of Advertising Research, been validated. 28 (1988): 11-31. Reynolds, T.J. and J.M. Phillips. “A Review and Comparative REFERENCES Analysis of Laddering `Research Methods: Recommendations for Quality Metrics.” In Review of Marketing Research, N. Ariely, D. and G.S. Berns. “Neuromarketing: the hope and hype of Malhotra (ed) (2009). neuroimaging in business.” Nature Reviews, 11 (2010): 284- Reynolds, T.J. and J. Rochon. “Strategy-Based Advertising 292. Research: Copy Testing is not Strategy Assessment.” Journal Bhatt, M.A. “Evaluation and associations: A neural-network model of Business Research, 22 (1991): 131-142. of advertising and consumer choice.” Journal of Economic Smidts, A., M. Hsu, A.G. Sanfey, M.A.S. Boksem, R.B. Ebstein, Behavior & Organization, 82(2012): 236-355 S.A. Huettel, J.W. Kable, U.R. Karmarkar, S. Kitayama, B. Briesch, R. and P. Rajagopal. “Neural Network Applications in Knutson, I.l Liberzon, T. Lohrenz, M. Stallen, and C. Yoon, Consumer Behavior.” Journal of Consumer Psychology, 20 Marketing Letters, 25 (2014): 257-267. (2010): 381-389 Advances in Consumer Research (Volume 45) / 847

West, P.M., P.L. Brockett, and L.L. Golden. “A Comparative Yoon C., R. Gonzalez, A. Bechara, G.S. Berns, A.A. Dagher, L. Analysis of Neural Networks and Statistical Methods for Dubé, S.A. Huettel, J.W. Kable, I. Liberzon, H. Plassmann, A. Predicting Consumer Choice.” Marketing Science, 16, 4 Smidts, and C, Spence. “Decision neuroscience and consumer (1997): 370-391. decision making.” Marketing Letters, 23 (2012): 473–485. The Intent to Persuade: Spontaneous Emotionality in Word-of-Mouth Communications Matthew D. Rocklage, Northwestern University, USA Derek D. Rucker, Northwestern University, USA Loran F. Nordgren, Northwestern University, USA

EXTENDED ABSTRACT ity. Moreover, it has been theorized that using emotionality to af- Consumers provide online product reviews for reasons that fect others can often be enacted without much deliberation due to range from a desire to help others to an effort to boost their own ego its relatively overlearned association with impacting others’ behav- (e.g., Hennig-Thurau et al. 2004). One prominent self-serving mo- iors (Frijda and Mesquita 1994). Therefore, we also manipulated tive among consumers is an intent to persuade others (Berger 2014). the cognitive load participants experienced. To begin, we ensured For example, individuals who want a restaurant to succeed might that all participants had equal knowledge of the product they would post and share a review to persuade others to dine there. Despite the write about. To this end, we had participants recall a novel they had recognition that people can be driven by the intent to persuade, far read that they would issue 5 stars. Next, we manipulated cognitive less research examines how this motive transforms how people com- load via a secondary task. Finally, we manipulated the intent to per- municate their views to others. To address this gap, we propose that suade more naturalistically via a referral program. Specifically, we the intent to persuade leads to subtle but detectable shifts in consum- informed half of the participants they would be paid $1.00 for each ers’ emotionality. Moreover, we demonstrate people turn to emotion- future participant who selected their book based on their review. We ality even when such appeals can backfire. replicated the results from Experiment 1: individuals with the intent to persuade expressed greater emotionality than those in the control The Intent to Persuade: A Catalyst for Emotionality condition. Moreover, these results were not moderated by cognitive Influential models of emotion posit that emotions evolved not load; even under high load individuals used greater emotionality. only to provide information to the person experiencing the emotion The pervasiveness of these effects raise the possibility that con- (Pham 2004), but also to influence others (Van Kleef 2009; Frijda sumers may rely on emotional arguments even in situations when and Mesquita 1994). Although research shows that emotional ap- such appeals may be suboptimal. Previous work indicates rational peals from others can affect participants’ behavior (e.g., Van Kleef, appeals are more impactful on audiences with cognitively-based atti- De Dreu, and Manstead 2004; Clark and Taraban 1991), there is a tudes (Fabrigar and Petty 1999) and that emotional appeals can even relative lack of research demonstrating whether and when individu- backfire with cognitive audiences (Haddock et al. 2008). Will people als themselves spontaneously increase their emotionality to impact continue to use more emotional appeals even when emotional ap- others. In the present work, we propose and explore the idea that the peals could backfire? Experiment 3 tested this possibility. intent to persuade acts as a catalyst for consumers to spontaneously In Experiment 3, we used a similar procedure to Experiment 2 engage in the use of emotionality. but we added conditions to test the effect of audience. Based on the An obstacle to answering this question for consumer commu- results from a pilot study, we conducted a preregistered experiment nications is that many communications in the modern world occur (https://osf.io/vbuqn/). There were four conditions in the experi- primarily through text (e.g., online reviews), thereby making the ment (n = 781). Two of the conditions (control and “Persuade – No measurement of emotionality difficult. However, Rocklage and Fazio Group”) provided similar instructions as in Experiment 2. However, (2015, 2016) have recently developed the Evaluative Lexicon (EL) two additional “Persuade” conditions were given specific groups to – a computational linguistic tool that enables researchers to quan- persuade. Based on pretesting, those in the “Emotionalists” condition tify language in terms of its implied emotionality. The EL allows re- were told their reviews would be shared with artists, dancers, and searchers to quantify differences between more emotional adjectives musicians from a group named “The Emotionalists.” Those in the such as “wonderful” and “lovable” versus more cognitive adjectives “Rationalists” condition were told they would be shared with a group such as “helpful” and “excellent.” The EL has been validated through of scientists, mathematicians, and economic analysts named “The in-laboratory experiments as well as in natural, archival text. Society for Applied Rationality and Mathematics.” We replicated the results of Experiment 2: those incentivized to persuade used greater The Current Research emotionality than the control condition. Most importantly, we found Experiment 1 examined the link between the intent to persuade that those in the “Rationalists” condition also used greater emo- and emotionality. Participants wrote a review for one of 20 products tionality compared to those in the control condition. Thus, despite (n = 778). To examine the extent of individuals’ use of emotional- previous research demonstrating the effectiveness of more rational, ity in persuasion, we utilized topics that varied both in emotionality unemotional appeals for such individuals, participants continued to (hedonic versus utilitarian products) as well as content (books versus utilize emotionality. non-book products). All participants wrote a 5-star review. However, those in the “Persuade” condition were given the additional instruc- Discussion tions to persuade others to purchase the product. Moreover, we com- Although prior work has enumerated the motives consumers pared the emotionality of these reviews to real-world 5-star reviews have for sharing information, little work has examined the conse- of the same products from Amazon.com (n = 840). These reviews quences of these motives. We find that the intent to persuade led con- provide a naturalistic baseline for the emotionality expressed toward sumers to spontaneously intensify the emotionality of their language. these products. Across the 20 products, participants with the intent This result occurred across topics and even in situations where emo- to persuade used greater emotionality compared to both those in the tional appeals might backfire. We also provide evidence that using control condition as well as the real-world Amazon.com reviews. emotionality may represent a default approach to persuasion that re- Experiment 2 (n = 288) was designed to provide a further test quires relatively few cognitive resources to implement. of the relationship between the intent to persuade and emotional-

Advances in Consumer Research 848 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 849 REFERENCES Hennig-Thurau, Thorsten, Kevin P. Gwinner, Gianfranco Walsh, Clark, Margaret S and Carolyn Taraban (1991), “Reactions to and and Dwayne D. Gremler (2004), “Electronic word-of-mouth willingness to express emotion in communal and exchange via consumer-opinion platforms: What motivates consumers to relationships,” Journal of Experimental Social Psychology, 27 articulate themselves on the Internet?,” Journal of Interactive (4), 324–36. Marketing, 18 (1), 38–52. Fabrigar, Leandre R. and Richard E. Petty (1999), “The role of the Pham, Michel T. (2004), “The logic of feeling,” Journal of affective and cognitive bases of attitudes in susceptibility to Consumer Psychology, 14 (4), 360–69. affectively and cognitively based persuasion,”Personality and Rocklage, Matthew D. and Russell H. Fazio (2015), “The Social Psychology Bulletin, 25 (3), 363–81. Evaluative Lexicon: Adjective use as a means of assessing and Frijda, Nico H. and Batja Mesquita (1994), “The social roles and distinguishing attitude valence, extremity, and emotionality,” functions of emotions,” in Emotion and culture: Empirical Journal of Experimental Social Psychology, 56, 214–27. studies of mutual influence, S. Kitayama and H. R. Markus, ——— and ——— (2016), “On the dominance of attitude eds., Washington, DC, US: American Psychological emotionality,” Personality and Social Psychology Bulletin, 42 Association, 51–87. (2), 259–70. Haddock, Geoffrey, Gregory R. Maio, Karin Arnold, and Thomas Van Kleef, Gerben A. (2009), “How emotions regulate social life: Huskinson (2008), “Should persuasion be affective or The Emotions as Social Information (EASI) model,” Current cognitive? The moderating effects of Need for Affect and Need Directions in Psychological Science, 18 (3), 184–88. for Cognition,” Personality and Social Psychology Bulletin, 34 ———, Carsten K. W. De Dreu, and Antony S. R. Manstead (6), 769–78. (2004), “The interpersonal effects of anger and happiness in negotiations,” Journal of Personality and Social Psychology, 86 (1), 57–76. Keep Calm and Smell the Roses: The Differential Impact of Low and High Arousal Emotions on Consumption Outcomes Maria Rodas, University of Minnesota, USA Rohini Ahluwalia, University of Minnesota, USA

EXTENDED ABSTRACT willingness to pay. Additionally, this type of focus on the consumer Consumption experiences can elicit different emotions (Cohen experience will slow down hedonic adaptation (i.e., the rate at which & Areni, 1991). For example, a consumer can feel energized while consumption happiness erodes). consuming a cold, fizzy drink, or calm while consuming a warm cup In four experimental studies, we find evidence that low (vs. of tea. While there has been research that explores the impact of the high) arousal consumption emotions will have a positive impact on consumption emotions’ level of arousal, most of this research has product liking (studies 1 and 2), willingness to pay (study 2), desire focused on high arousal emotions. For example, high arousal emo- to continue the consumption experience (study 3), and self-brand tions have been found to have a positive effect on information shar- connection (studies 2 and 4). We also find that low arousal consump- ing and talking (Berger, 2011), as well as on desire and readiness tion emotions slow down hedonic adaptation (i.e., the rate at which for action (Rucker & Petty, 2004). And while consumption can also the happiness derived from the consumption erodes) in a longitudinal elicit low arousal emotions (Richins, 1997), they have received less study with actual purchases (study 4). Furthermore, we find evidence examination. In our research we address this gap in the literature by that the level of emotional arousal impacts the extent to which peo- focusing on low arousal emotions (e.g., contentment, peacefulness, ple’s mind wanders off from the consumption experiences (as indi- relaxation) and identifying their impact on several favorable con- cated by participants in study 1) and the extent to which they remem- sumption outcomes. Specifically, we propose and find evidence in ber details of the consumption experience (specifically, accurately four studies that experiencing low (vs. high) arousal emotions during remembering the name of a slow or fast tempo song in study 2). Our consumption has a positive impact on product liking, willingness to studies span different types of products consumed (pen, song, tea, pay, self-brand connection, and desire to continue the consumption actual purchases), different types of manipulation of level of arousal experience, as well as slows down hedonic adaptation (or the rate at (incidental, stimulus-generated, primed with product positioning and which utility derived from the consumption erodes). We find that this communications), and different methodologies (consumption in lab, occurs because low arousal emotions, during a consumption experi- longitudinal study), adding robustness to our findings. ence, reduce the extent to which the mind wanders to other thoughts and stimuli, increasing the consumer’s focus on the consumption REFERENCES experience itself, allowing the consumer to develop stronger connec- Barrett, L. F., & Russell, J. A. (1999). The structure of current tions with the product. affect: Controversies and emerging consensus. Current We follow past classifications of emotions (Barrett & Russell, Directions in Psychological Science, 8(1), 10-14. 1999; Mogilner, Aaker, & Kamvar, 2012; Tsai, Knutson, & Fung, Berger, J. (2011). Arousal increases social transmission of 2006) in using level of arousal as a key dimension for categorizing information. Psychological Science, 22(7), 891-893. emotions. Low arousal emotions include emotions such as peaceful, Cohen, J. B., & Areni, C. S. (1991). Affect and Consumer Behavior. calm, and relaxed, whereas high arousal emotions include emotions In T. S. Robertson & H. H. Kassarjian (Eds.), Handbook of such as excited, elated, and energized. We propose that high arousal Consumer Behavior (pp. 188-240). Englewood Cliffs, NJ: (vs. low arousal) accelerates the speed of people’s thoughts. Past re- Prentice-Happ. search has found that thought speed creates an urge for action that has MacKenzie, S. B. (1986). The role of attention in mediating the been compared to that of fight-or-flight (Pronin, 2013). When people effect of advertising on attribute importance.Journal of experience this type of response, their attention narrows (Skosnik, Consumer Research, 13(2), 174-195. Chatterton, Swisher, & Park, 2000). Low arousal emotions, on the Mogilner, C., Aaker, J., & Kamvar, S. D. (2012). How happiness other hand, are likely to slow down the speed of people’s thoughts, affects choice.Journal of Consumer Research, 39(2), 429-443. broadening their focus of attention. Pronin, E. (2013). When the mind races: Effects of thought speed Based on these findings, we hypothesize that low arousal posi- on feeling and action. Current Directions in Psychological tive emotions (e.g., contentment, serenity, calmness) will slow down Science, 22(4), 283-288. the speed of both actions and thoughts. The reduced action orienta- Richins, M. L. (1997). Measuring emotions in the consumption tion and increased tendency to talk and do less (which comes with experience. Journal of Consumer Research, 24(2), 127-146. quietude and contentment), slows down the mind from jumping to Rucker, D. D., & Petty, R. E. (2004). Emotion specificity and other thoughts and stimuli, instead allowing it to focus on the current consumer behavior: Anger, sadness, and preference for experience and increasing the extent to which the consumer absorbs activity. Motivation and Emotion, 28(1), 3-21. it in, experiences it more fully, and thinks about it. This increased Sheldon, K. M., & Lyubomirsky, S. (2012). The challenge of focus on the experience can have several benefits. Past research has staying happier: Testing the hedonic adaptation prevention shown that merely giving more attention to a target can imbue it with model. Personality and Social Psychology Bulletin, 38(5), more value and weight (MacKenzie, 1986). Furthermore, Sheldon & 670-680. Lyubomirsky’s (2012) hedonic adaptation prevention model posits Skosnik, P. D., Chatterton, R. T., Swisher, T., & Park, S. (2000). the more people keep thinking about the target, the more they contin- Modulation of attentional inhibition by norepinephrine and ue to derive positive feeling from it, sustaining their initial happiness cortisol after psychological stress. International Journal of boost. Thus, we propose low arousal emotions, by allowing people Psychophysiology, 36(1), 59-68. to focus on the consumption experience, will result in consumers Tsai, J. L., Knutson, B., & Fung, H. H. (2006). Cultural variation in feeling a stronger sense of connection with the consumed product, affect valuation.Journal of Personality and Social Psychology, which will in turn have other positive consequences, such as higher 90(2), 288.

Advances in Consumer Research 850 Volume 45, ©2017 Thicker than Water: The Influence of Familism on Consumer Response to Brand Extensions Maria A. Rodas, University of Minnesota, USA Michael J. Barone, University of Louisville, USA Carlos J. Torelli, University of Illinois at Urbana-Champaign, USA

EXTENDED ABSTRACT the extension is too far from the brand (e.g., a calculator as an athletics Brand extension strategies are widely used by firms to leverage brand extension), the similarities between the concepts of family and the equity built into their brands. Every year, companies spend billions brands will weaken, which will prevent the family schema to be ap- of dollars introducing new products under established brand names, as plied. exemplified by Apple’s iPhone, Arm & Hammer toothpaste, and Jeep Across three experimental studies, we show that familism posi- baby strollers. Despite firms’ reliance and investment on this strategy, tively impacts consumers’ response to brand extensions that don’t many brand extensions fail. In this regard, research has identified a stray too far from the parent brand and which are lower in price and number of key determinants that shape brand extension success. For quality than the parent brand. We observe this effect when familism is instance, consumer acceptance of brand extensions may be shaped by measured as a chronic individual difference (study 1), and when it’s their trust toward or liking of the parent brand as well as the perceived primed with writing exercised (studies 2 and 3). We also find evidence fit between the parent brand and extension (see Loken, Joiner, & Hous- in studies 1 and 3 that the underlying mechanism is an increase in ton, 2010, for review). Additionally, the extent to which brand exten- perceived similarity with the parent brand, providing evidence that sions are viewed favorably can be determined by consumers’ risk aver- family schemas are being mapped onto the brand extension domain. sion (Yeo & Park, 2006), innovativeness (Völckner & Sattler, 2006), We identify distance from the parent brand as a boundary condition and cultural orientation (e.g., Ng & Houston, 2006). (studies 1 and 2), rule out analytic-holistic thinking as an alternative We explore a novel consumer characteristic that impacts consum- explanation (study 1), and provide evidence that the effects of familism ers’ responses to brand extensions, familism. Familism is defined as a extend beyond the nuclear family. Furthermore, our studies span dif- strong identification and attachment of individuals with their families ferent populations (college students, MTurkers) and different catego- (nuclear and extended) and strong feelings of loyalty, reciprocity and ries (electronics, athletics), adding robustness to our findings. solidarity among members of the same family (Triandis, Marin, Betan- This research aims to contribute to the brand extension litera- court, Lisansky, & Chang, 1982). We propose and find evidence that ture by empirically investigating a novel consumer characteristic that familism (either chronic or primed) can improve consumers’ evalua- impacts consumers’ responses to brand extensions, namely familism. tions of brand extensions by increasing perceptions of the extensions Furthermore, our work has important managerial implications. Every fit with the parent brand. We theorize that this is the case because year companies spend billions of dollars introducing new brand exten- familism makes the family schema more accessible, which activates sions, and many fail. Our research could provide guidance to mini- associations of relatedness. mize the risk of failure. For example, if a brand wants to downwardly Terminology used in past literature hints at the application of a extend itself by introducing a brand extension that is lower in price/ family schema when discussing brand extensions. For example, this quality, it could consider targeting consumer who are chronically high literature constantly refers to a “parent brand” to indicate the brand in familism (e.g., Hispanics) or it could prime familism in its commu- from which the extension originates, a “brand family” in reference to nications (e.g., by portraying families). the group of products and extensions that fall under a given brand, and a “brand offspring” to indicate a brand extension. There is also evi- REFERENCES dence that consumers use relationships, including family, as metaphors Fournier, S. (1998). Consumers and their brands: Developing when building bonds with brands (Fournier, 1998). However, the ques- relationship theory in consumer research. Journal of Consumer tion remains whether consumers use the family schema to guide their Research, 24(4), 343-373. attitudes when evaluating brand extensions. We propose that it does, Heath, T. B., & Chatterjee, S. (1995). Asymmetric decoy effects on resulting in a perception of greater similarity between the parent brand lower-quality versus higher-quality brands: Meta-analytic and and offspring, which in turn positively impacts the evaluations of cer- experimental evidence. Journal of Consumer Research, 22(3), tain brand extensions. 268-284. Generally speaking, people value family. However, what distin- Loken, B., Joiner, C., & Houston, M. J. (2010). Leveraging a brand guishes people high in familism is that they tend to put the needs of through brand extension: A review of two decades of research. their family above their own. This type of ties to family is particularly In B. Loken, R. Ahluwalia & M. J. Houston (Eds.), Brands and apparent when family members are in need, or not doing very well. As brand management: Contemporary research perspectives (pp. such, we propose that the effect of familism will only emerge when 11-41). the brand extension is of lower price and quality than the parent brand. Ng, S., & Houston, M. J. (2006). Exemplars or beliefs? The impact Another reason why we don’t expect to see any differences with high of self-view on the nature and relative influence of brand price/quality brand extensions is related to the asymmetric effect of associations. Journal of Consumer Research, 32(4), 519-529. quality (Heath & Chatterjee, 1995). When the price and quality of a Triandis, H. C., Marin, G., Betancourt, H., Lisansky, J., & Chang, given product is good, people can appreciate that. It’s only when the B.-H. (1982). Dimensions of familism among Hispanic and price/quality is bad that people use other heuristics (e.g., decoys, sche- mainstream Navy recruits: DTIC Document. mas) to make their evaluations. Völckner, F., & Sattler, H. (2006). Drivers of brand extension Furthermore, we propose the distance from the parent brand as a success. Journal of Marketing, 70(2), 18-34. boundary condition. More specifically, we propose and find evidence Yeo, J., & Park, J. (2006). Effects of parent-extension similarity that the effect of familism only emerges when evaluating brand exten- and self regulatory focus on evaluations of brand extensions. sions that don’t stray too far from the parent brand. We argue that when Journal of Consumer Psychology, 16(3), 272-282.

Advances in Consumer Research 851 Volume 45, ©2017 Preference Refinement After a Budget Contraction Gretchen Ross, Penn State University, USA Kurt A. Carlson, Georgetown University, USA Margaret G. Meloy, Penn State University, USA

EXTENDED ABSTRACT p=.003. There was an additional decline between the second 21 day Most models of consumer choice assume that preferences are allocation on the first day and the 21 day allocation two days later, independent of budget levels. Specifically, consumers are thought though it was ns. Preference refinement that began as a result of to react symmetrically to increasing and decreasing budgets, landing contraction out-lasted desire to make allocations that were consistent at their optimal combination of options for any given budget level. with initial allocations. Empirically, however, prior research has demonstrated that this is not the case (Carlson, Wolfe, Blanchard, Huber and Ariely 2015; Dargay Study 3: Strength of Contraction 2001; Kamakura and Du 2012; Shea 1995). For example, Carlson Study 3 examined different levels of budget contractions (21- et al. (2015) showed that rather than making small changes in con- 7-21 vs 21-14-21) in the domain of time. Those experiencing a sumption across all options, consumers who faced a contracting bud- more extreme contraction face more trade-offs, which should lead get cut out entire categories of consumption while maintaining full to greater preference refinement. Students (N=200, 49% female; me- consumption in others. dian age=19) exhibited greater preference refinement under an ex-

In the current work, we examine whether allocations of a given treme contraction (21-7-21: MD=0.64 (SD=2.24, t(99)=2.86, p=.005) budget for a person are the same before versus after a budget contrac- than under a modest contraction (21-14-21: MD=0.28 (SD=1.62, tion (i.e., when the budget re-expands to pre-contractionary levels). t(99)=1.73, p=.09). Extreme contractions led to greater preference In other words, do the allocations that emerge under budget contrac- refinement. tions reflect a temporary shift in preferences akin to transient preference shifts, or are these shifts in preference more stable and lasting? Study 4: Trade-Offs and Dispelling Anchoring We contend that the trade-offs necessary during a contractionary pe- Trade-offs made during the contractionary period should medi- riod persist when the budget re-expands; consumers engage in a pro- ate the relationship between condition and final allocation. To exam- cess of refining their preferences as they reconcile trade-offs (Hoef- ine this, MTurk workers (N=420, 42% female; median age=36) were fler and Ariely 1999) and figure out what really matters to them. We randomly assigned to either simple expansion ($350-$850) or a con- use change in number of unique items selected as our metric of pref- traction re-expansion condition ($850-$350-$850) (using the stimuli erence refinement. That is, we count the number of unique items to from S1B). Three measures of trade-offs were completed after the $350 allocation: “Were you making trade-offs?” (1=“definitely not” which the re-expanded budget was allocated (X2) and difference it from the initial budget allocation (X ) (M = X -X ). We then test for to 7=“definitely yes”); “Were your decisions governed by trade- 1 D 1 2 offs?” (1=“not at all” to 7 =“extremely”); and “How many trade-offs a positive difference (MD>0) to determine if there has been a refinement in preferences. did you make during the $350 allocation?” (slider bar from 0 to 10). For simple expansion, allocations moved from 7.16 to 8.23 Studies 1A-1C: Time, Money and Space as the budget increased. In the contraction re-expansion condition, Study 1 examines the preference refinement effect across time, we observed the standard preference refinement effect (MD=0.54, money and space. Specifically, study 1A (N=119 students, 43% fe- SD=2.01, t(210)=3.87, p<.001) and directionally less variety in the male, median age=19) examined allocations of a sequence of travel final allocation under re-expansion (7.84) than under simple expan- budgets (21 days to 7 days to 21 days) to cities. Study 1B (N=130 sion (8.23). This result helps dispel anchoring as an alternative expla- MTurkers, 42% female; median age=36) examined allocations of nation for the findings. Additionally, individuals believed trade-offs money to holiday gift recipients under a sequence of financial bud- were more instrumental under re-expansion compared to simple ex- gets ($850-$350-$850). Study 1C (N=276 students, 30% female; pansion (all p<.01). Trade-offs partially mediated the relationship be- median age=19) measured allocations of vegetable seeds across a tween (re)expansion and the final $850 allocation (PROCESS, model budget of planting rows in a garden plot (21-10-21). In all studies, 4, b=0.30, SE = 0.08, BC 95% CI = [0.16, 0.47]. participants exhibited preference refinement (Study1A: M =0.42, D Study 5: Self- Versus Other-Determined Allocation SD=1.49, t(118)=3.07, p=.003; Study 1B: MD=0.31, SD=1.10), Variety t(129)=3.19, p=.002; Study 1C: MD=0.18, SD=1.18, t(275)= 2.51, p=.01). Participants consistently allocated their final budget to fewer Study 5 examines the source of constrained choice in the con- unique items than they allocated the same budget before experienc- traction period as either self- or other-determined. If the contraction- ing a contraction (i.e., preference refinement occurred in all three ary choice is made by another, participants will not need to make studies). trade-offs, and we should not observe preference refinement. Stu- dents (N=219, 46% female; median age=18) participated. In the self- Study 2: Time Delay determined condition, the 7-day allocation was based on personal Study 2 was conducted to examine the opposing forces of pref- preference. In the other-determined condition, a travel partner picked erence refinement and preference consistency by examining alloca- three cities for the 7-day trip. Participants completed trade-off questions two days apart. Student participants (N=77, 56% female; me- tions. dian age=19) allocated time to travel (as in S1A); 63 completed a When self-determined (N=113), preference refinement occurred follow-up 21 day allocation two days later. Preference refinement ap- (MD=0.36, SD=1.76, t(112)=2.19, p=.03). However, when the contraction allocation was driven by another’s preferences (N=106), peared in the initial 21-7-21 survey (MD=0.33, SD=1.15), t(62)=2.30, p=.03. Two days later, the 21 day allocation was significantly lower the preference refinement effect did not obtain (MD=-0.22, SD=1.13, than the initial 21 day allocation (MD=0.46, SD=1.20), t(62)=3.04,

Advances in Consumer Research 852 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 853 t(105)=1.98, p=.05). If anything, participants exhibited preference Dargay, Joyce M. (2001), “The Effect of Income on Car Ownership: expansion under this condition. Evidence of Asymmetry,” Transportation Research Part A: Policy and Practice, 35 (9), 807-21. Discussion Hoeffler, Steve and Dan Ariely (1999), “Constructing Stable When budgets contract, consumers face difficult trade-offs that Preferences: A Look Into Dimensions of Experience and help them discover what really matters to them. This process leads Their Impact on Preference Stability,” Journal of Consumer to non-transient preference refinement – preferences were narrower Psychology, 8 (2), 113-39. and more consistent after a contraction. The work has implications Kamakura, Wagner A. and Rex Yuxing Du (2012), “How Economic for brand relationships that must be reinforced during economic Contractions and Expansions Affect Expenditure Patterns,” downturns to prevent less-preferred brands from being cut perma- Journal of Consumer Research, 39 (2), 229-47. nently. Shea, John (1995), “Myopia, Liquidity Constraints, and Aggregate Consumption: A Simple Test,” Journal of Money, Credit and REFERENCES Banking, 27 (3), 798-805. Carlson, Kurt A., Jared Wolfe, Simon J. Blanchard, Joel C. Huber, and Dan Ariely (2015), “The Budget Contraction Effect: How Contracting Budgets Lead to Less Varied Choice,” Journal of Marketing Research, 52 (June), 337-48. How Discount Price Negatively Influences Ethical Consumption Yuhosua Ryoo, University of Texas at Austin, USA Minette Drumwright, University of Texas at Austin, USA

EXTENDED ABSTRACT questions concerning the perceived product efficacy with three items Over the years, the topic of pricing has been of great inter- on a seven point scale (White, MacDonnell, and Ellard, 2012). Social est in ethical consumption. In particular, the high price of ethical desirability (Ballard 1992) was also measured and controlled as a products has been considered an obstacle for consumers when they confounding variable. The PROCESS analysis (Hayes 2013, Model want to make an ethical purchase decision. To promote consumers’ 8, 5000 bootstrap resamples) showed a significant mediated modera- purchase of ethical products, marketers and scholars have suggested tion effect (95% CI, -.39 to -.13). In particular, the price × ethical that ethical products need to be offered at a discount price. The cur- mindset interaction had a significant effect on perceived product ef- rent research, however, made a counterargument that a discount price ficacy (β = -.32, t = -4.32, p < .001; 95% CI, -.43 to -.17). Perceived negatively influences consumers’ purchase behavior. Guided by the product efficacy, in turn, had a significant positive effect on WTPT-shirt price-product efficacy beliefs, we argued that consumers perceive (β = .752, t = 7.965, p < .001; 95% CI, .56 to .94). Consistent with ethical products with discount prices to be less efficient, which in the predictions, participants with high levels of ethical mindset in the turn leads them less willing to purchase the discounted ethical prod- discounted price condition (M = 4.83) were less likely to purchase ucts. the ethical T-shirt than those in the premium price condition (M = To demonstrate the hypothesis, four experiments were conduct- 6.32; 95% CI, -1.01 to -.49) when the perceived product efficacy was ed. In Experiment 1, 75 participants who were recruited from Ama- included in the model. Participants’ willingness to purchase an ethi- zon’s MTurk randomly assigned either to the discount price condi- cal T shirt was not different among those with low levels of ethical tion or to the premium price condition. Whereas participants in the mindset (Mpremium = 5.37 vs. Mdiscount = 5.3; 95% CI, -.32 to .16). discount price condition received information that organic tomatoes In the last experiment, we suggested how additional informa- are on sale for a dramatically discount price ($1.89 → $ 0.89/lb), tion about product efficacy can prevent the boomerang effects of those in the premium price condition were told that organic tomatoes discount prices. The procedure and measurements were in keeping are selling at a premium price ($1.89) as usual. Participants then ex- with those used in Experiment 4. However, in Experiment 4, 117 pressed how they are willing to purchase organic tomatoes with five participants were exposed only to a discount price even if the half of items on a seven point scale (Dodds, Monroe, and Grewal 1991). The them received information that boosts the efficacy of the discounted findings provided preliminary evidence that organic foods with dis- ethical products whereas the others did not. The significant addi- count prices (M = 4.65) can decrease consumers’ willingness to pur- tional information × ethical mindset interaction (β = -1.25, p < .001) chase the organic foods (vs. premium price: M = 6.03; F = 18.052, indicated that participants did not receive information about product p < .001). efficacy were less likely to purchase the discounted fair trade cof- The purpose of Experiment 2 was to demonstrate the moderat- fee coupons as their ethical mindset (EMCB) increased (β = -.91, p ing role of consumers’ ethical mindsets. A total of 125 participants < .05). However, when the information was presented, the level of were recruited from Amazon’s MTurk and randomly assigned either ethical mindset (EMCB) did not predict the likelihood of purchasing to the discount price condition or to the premium price condition. the discounted coupon (β = .25, p > .1) Participants were shown an online coupon for a cup of fair trade The findings collectively indicated that price promotions with- coffee that is available any grocery store near them. In the discount out a consideration of consumers’ belief system would backfire. The price condition, the coupon was worth $0.89, whereas the coupon in present research first demonstrated the boomerang effect of ethical the premium price condition was worth $2.56. After expressing their products with discounted prices in the wide range of the ethical con- willingness to purchase the fair trade coffee coupon, participants also sumption contexts (e.g., fair trade coffee, organic foods, and ethical measured their moods with the PANAS scale (Watson, Clark, and clothes). In addition, the present research provided deeper under- Tellegen 1988) and ethical mindset with the EMCB scale (Sudbury- standings about consumers’ responses to pricing actions by demon- Riley and Kohlbacher 2016). After controlling for the mood effects, strating a mediating role of the product efficacy. In particular, this the results showed that there was a significant interaction effect be- research suggested a boundary condition of the boomerang effects tween the price of a fair trade coffee coupon and consumers’ mindset of discount prices by examining the moderating role of consumers’ (β = .271, t = 3.957, p < .001). The spotlight analysis (Aiken and ethical mindsets. This highlighted that the individual-related factor West 1991) revealed that the boomerang effects of discount prices plays a vital role in ethical consumption and the boomerang effects are predominant in certain consumer segments. are prevalent among those have high levels of ethical mindset (Mdis- count = 4.97 vs. Mpremium = 6.54; t = -5.82, p < .001), but this relationship was not significant among those with low levels of ethical mindset REFERENCES Aiken, Leona S., Stephen G. West, and Raymond R. Reno (1991), (Mdiscount = 6.3 vs. Mpremium = 6.39; p > .1). In the context of the ethical clothes consumption, Experiment Multiple Regression: Testing and Interpreting Interactions. 3 aimed to demonstrate how the perceived product efficacy mediates Newbury Park, CA: Sage. a negative impact of discount prices. A total of 135 participants who Ballard, Rebecca (1992), “Short Forms of the Marlowe-Crowne were recruited from Amazon’s MTurk were randomly assigned to a Social Desirability Scale,” Psychological Reports, 71 (3), price condition (premium vs. discounted). They were shown a picture 1155-1160. of an ethical T-shirt and hypothetically given a chance to purchase Dodds, William B., Kent B. Monroe, and Dhruv Grewal (1991), the product. The price was manipulated by indicating different prices “Effects of Price, Brand, and Store Information on Buyers’ on the label of the T-shirt (i.e., $2.5 for the discount price condition Product Evaluations,” Journal of Marketing Research, 307- vs. $5 for the premium price condition). Along with the questions 319. that were measured in Experiment 2, participants then answered

Hayes, Andrew F (2013), Introduction to Mediation, Moderation, White, Katherine, Rhiannon MacDonnell, and John H. Ellard and Conditional Process Analysis: A Regression-Based (2012), “Belief in a Just World: Consumer Intentions and Approach. NY: Guilford Press. Behaviors Toward Ethical Products,” Journal of Marketing, 76 Sudbury-Riley, Lynn and Florian Kohlbacher (2016), “Ethically (1), 103-118. Minded Consumer Behavior: Scale Review, Development, and Validation,” Journal of Business Research, 69 (8), 2697-2710. Watson, David, Lee A. Clark, and Auke Tellegen (1988), “Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS Scales,” Journal of Personality and Social Psychology, 54 (6), 1063. Driven by the Cues: Goal Attainment Makes Consumers More Responsive to Shopping-Related Contextual Cues Sebastian Sadowski, Maastricht University, The Netherlands Bob Fennis, University of Groningen, The Netherlands Koert van Ittersum, University of Groningen, The Netherlands

EXTENDED ABSTRACT Study 1 The literature on consumer behavior considers shopping be- 100 students participated in a 2 (goal state: anticipation vs. at- havior as predominantly goal-oriented (e.g., Kopetz et al., 2012). tainment) × 2 (assortment structure: equivalent vs. nonequivalent) Nonetheless, shopping goals do not always need to be at the top of mixed design experiment, where goal state was manipulated be- consumers minds and they can sometimes be pushed to the back- tween-subjects and assortment structure within-subjects. ground by other, overriding, more desired goals. To date, scant re- Participants were invited to play a Tic-Tac-Toe game and were search has been devoted to the interplay between consumer shopping informed that the study examined their performance during games. goals and other salient, concomitantly active goals and the impact of Five games of Tic-Tac-Toe were played in total by each participant. such interference on consumer decision-making. Various goals can We varied goal states by either asking participants to make the choice accompany consumers in their daily life (e.g., taking a train, arriving of their most preferred product before (anticipation) or after (attain- at a specific location, going to work) and disparate goals of differ- ment) they have played the game. Participants selected products ing motivational intensity could be active at the same time (Pieters, from two different assortments: evidently equivalent (Ice Creams) Baumgartner and Allen, 1995). In this research we focus particularly and nonequivalent (Candy Packages) assortments, consisting of 64 on the anticipation and the attainment of such highly desired goals, different products.Our focal dependent variable was the centrality of investigating how either having a salient, highly desired goal that the product choice, coded from 1 to 7, with the higher value repre- needs to be attained in the near future or having attained this par- senting a more central option chosen. ticular goal influences responsiveness to shopping-related contextual Results cues and subsequently location-based preferences, that is, selecting A repeated measures ANOVA with goal state as a between- products from specific locations. subjects factor and assortment structure as a within-subjects factor We posit that when a shopping goal is pushed to the background, revealed a significant interaction between goal state and assortment the influence of shopping-related cues on consumer decision-making structure (F(1, 90)=4.04, p=.05). Additional simple main effects will be attenuated (Ferguson & Bargh, 2004). Therefore, when other analysis showed that the interaction was predominantly driven by salient, highly desired goals are potent enough to inhibit the goal of participants in the goal attainment condition. Participants who have shopping, we expect that consumers become less susceptible to the already attained monetary rewards and were selecting products from influence of shopping-related contextual cues (i.e., assortment struc- an equivalent assortment were more likely to go to the center of the ture). On the other hand, when other, shopping-unrelated goals have equivalent assortment (M=4.44, SD=1.53) in comparison with those already been attained, consumers open up and approach shopping in who attained monetary rewards and were choosing products from a a more goal-oriented way, dedicating greater processing capacity to nonequivalent assortment (M=3.42, SD=1.62; F(1,90)=9.86, p<.01). their product selection (Gable & Harmon-Jones, 2010). They process When the rewards were anticipated assortment structure did not steer not only separate products, but also their global visual configura- participants to choose products from different locations F( < 1). tion. As a result, consumer decision-making is more likely then to be shaped by subtle contextual manipulations in product organizations. Study 2 Our expectations are derived predominantly from goal systems 100 students participated in a 2 (goal state: anticipation vs. at- theory (Kruglanski et al., 2012). Various goals that are simultane- tainment) × 2 (assortment structure: equivalent vs. nonequivalent) ously activated compete for cognitive resources, pulling mental mixed design experiment, where goal state was manipulated be- resources from each other. The more resources are dedicated to a tween-subjects and assortment structure within-subjects. particular goal, the less resources are left for other coinciding goal Participants were approached with a survey either before they pursuits (Kruglanski et al., 2002). An important factor determining ate in a Burger King restaurant or after they had already eaten there. the allocation of resources across competing goals is motivational Participants were presented with two different assortments: evidently intensity. Goals characterized by highest motivational intensity pull equivalent (donuts) and nonequivalent (bread spreads). For each as- the most mental resources, leaving only a limited processing capacity sortment 99 products were shown (9×11 grid). for other, less desired but coinciding goals and associated with them Akin to study 1, our core dependent variable was the centrality contextual cues (Kruglanski et al., 2002). Therefore, we expect that of the choice, coded from 1 to 10, with the higher value representing sensitivity towards goal-irrelvant contextual cues is attenuated when a more central option chosen. other, more desired goals are concurrently active. In order to test this proposition, following Bar-Hillel (2015), we Results expect that when other goals have been attained, consumers will se- A repeated measures ANOVA with goal state as a between- lect products located closer to the center of an assortment (i.e., edge subjects factor and assortment structure as a within-subjects factor aversion) while selecting from equivalent assortments (assortments revealed a significant interaction between goal state and assortment composed of products varying only across a few attributes), whereas structure (F(1, 89)=5.07, p=.03). Additional simple main effects selecting products from nonequivalent assortments (assortments analysis showed that, in line with our expectations, the interaction composed of products varying across numerous attributes) will shift was predominantly driven by participants in the goal attainment consumers’ product preferences closer to the edge of an assortment condition. Participants who had already eaten and were selecting (i.e., edge preference). products from an equivalent assortment were more likely to go to the center of the equivalent assortment (M=6.54, SD=2.12) in com-

Advances in Consumer Research 856 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 857 parison with participants who had already eaten and were choosing Keller, C., Markert, F., & Bucher, T. (2015), “Nudging Product products from a nonequivalent assortment (M=5.41, SD=2.32; F(1, Choices: The Effect of Position Change on Snack Bar Choice,” 89)=10.67, p<.01). The structure of assortment did not make par- Food Quality and Preference, 41, 41-43. ticipants choose products from different locations when they were Kopetz, C. E., Kruglanski, A. W., Arens, Z. G., Etkin, J., & anticipating to satisfy their hunger (F < 1). Johnson, H. M. (2012), “The Dynamics of Consumer Behavior: A Goal Systemic Perspective,” Journal of Consumer REFERENCES Psychology, 22(2), 208-223. Bar-Hillel, M. (2015), “Position Effects in Choice from Kruglanski, A. W., Shah, J. Y., Fishbach, A., Friedman, R., & Chun, Simultaneous Displays a Conundrum Solved,” Perspectives on W. Y. (2002), “A Theory of Goal Systems,” In: MP Zanna Psychological Science, 10(4), 419-433. (Ed.), Advances in experimental social psychology. Baumgartner, H., Pieters, R., Haugtvedt, C., Herr, P., & Kardes, F. Phillips, L. W., & Sternthal, B. (1977), “Age Differences in (2008), “Goal-directed Consumer Behavior,” Handbook of Information Processing: a Perspective on the Aged Consumer,” consumer psychology. Journal of Marketing Research, 444-457. Ferguson, M. J., & Bargh, J. A. (2004), “Liking is for Doing: The Pieters, R., Baumgartner, H., & Allen, D. (1995), “A Means-End Effects of Goal Pursuit on Automatic Evaluation,”Journal of Chain Approach to Consumer Goal Structures,” International Personality and Social Psychology, 87(5), 557. Journal of Research in Marketing, 12(3), 227-244. Gable, P., & Harmon-Jones, E. (2010), “The Motivational Raghubir, P., & Valenzuela, A. (2006), “Center-of-Inattention: Dimensional Model of Affect: Implications for Breadth of Position Biases in Decision-Making,” Organizational Attention, Memory, and Cognitive Categorisation,” Cognition Behavior and Human Decision Processes, 99(1), 66-80. and Emotion, 24(2), 322-337. Salthouse, T. A. (1996), “The Processing-Speed Theory of Adult Harmon-Jones, E., Gable, P. A., & Price, T. F. (2013), “Does Age Differences in Cognition,”Psychological Review, 103(3), Negative Affect Always Narrow and Positive Affect Always 403. Broaden the Mind? Considering the Influence of Motivational Valenzuela, A., & Raghubir, P. (2009), “Position-Based Beliefs: Intensity on Cognitive Scope,” Current Directions in The Center-Stage Effect,”Journal of Consumer Psychology, Psychological Science, 22(4), 301-307. 19(2), 185-196. Iyengar, S. S., & Lepper, M. R. (2000), “When Choice is Demotivating: Can One Desire Too Much of a Good Thing?,” Journal of Personality and Social Psychology, 79(6), 995- 1006. 858 / Word of Mouth as Compensatory Consumer Behavior: Can Talking About Brands Restore Consumers’ Self-Concepts After Self-Threat? Word of Mouth as Compensatory Consumer Behavior: Can Talking About Brands Restore Consumers’ Self-Concepts After Self-Threat? Christina Saenger, Youngstown State University, USA Veronica L. Thomas, Towson University, USA Dora Bock, Auburn University, USA

EXTENDED ABSTRACT for the brand. Participants reported self-perceptions of intelligence Self-threat occurs when a person encounters information that (α = .79) using items adapted from Rick and Schweitzer (2012). Psy- calls into question the positivity of a given self-view (Gao et al. chological discomfort (α = .81) was measured with items adapted 2009), resulting in psychological discomfort (Festinger 1957). Sym- from Elliot and Devine (1994). bolic self-completion (Wicklund and Gollwitzer 1981) is one strategy A two-way ANOVA with threat and brand as IVs, covariate age, (Mandel et al. 2017) consumers deploy in response. As supported by and self-perceptions as DV resulted in no main effect for brand (p = the compensatory consumption literature, products can be consumed .50) or threat condition (p = .36), but the predicted threat by brand for psychological value (Ariely and Norton 2009) and their symbolic interaction emerged (F(1, 135) = 5.19, p < .05), replicating the same associations incorporated into consumers’ self-concepts (Belk 1988). pattern observed in study 1. Self-threat also affects consumers’ decision to spread WOM. PROCESS model 7 (Hayes 2013; 5,000 bootstrap resamples) People are more likely to spread WOM about symbolic products supports psychological discomfort as a mediator of the threat by (Chung and Darke 2006) that signal identity-related aspects of their brand interaction self-perceptions (index of moderated mediation: self-concepts (Berger 2014), and to spread WOM after self-threat .54, 95% CI = [.1038, 1.054]). The interaction affects psychologi- (Packard and Wooten 2013). We propose that WOM can actually re- cal discomfort (b = -1.19, t = -2.47, p < .05) and psychological dis- solve specific, self-concept related self-threats when spread about a comfort reduces self-perceptions (b = -.45, t = -9.68, p < .001). For brand that is symbolically congruent with the threat domain, termed the non-restorative brand, the indirect effect suggests psychological a restorative brand. Under self-threat, consumers who spread WOM discomfort explains threat’s negative effect on self-perceptions (-.57, about a restorative (non-restorative) brand should exhibit higher 95% CI = [-.9741, -.2211]), but was not significant for the restorative (lower) self-perceptions on the threatened attribute (H1). Further, brand (-.03, 95% CI = [-.3378, .2575]). Self-threat negatively affects psychological discomfort should mediate the interactive effect of self-perceptions through heightened psychological discomfort when threat and brand restorativeness on self-perceptions (H2). Evidence WOM is spread about a non-restorative brand, partially supporting suggests that self-esteem may moderate the proposed effects as H2. low-esteem consumers feel lower self-efficacy (Sherer et al. 1982) Study 3 mirrors study 1, except all participants (n = 224, aver- and exhibit stronger attachment to symbolic brands (Dommer et al. age age 38, 40% male) were exposed to self-threat and completed 2013). Under self-threat, spreading WOM about a restorative (non- three items measuring trait self-esteem (α = .87; Rosenberg 1965), restorative) brand will enhance (reduce) self-perceptions on the along with items measuring self-perceptions (α = .64) and psycho- threatened attribute through psychological discomfort for low self- logical discomfort (α = .82). esteem consumers (H3). PROCESS model 1 was significant (R2 = .17; F (3, 220) = 15.45, Study 1: Participants (n = 128, average age 36, 46% male) p < .001). Direct effects of brand (b = 1.20, t = 2.24, p < .05) and self- were randomly assigned to a condition where they wrote about a esteem (b = .41, t = 5.85, p < .001) were qualified by an interaction time they acted ethically (no self-threat) or unethically (self-threat). (b = -.19, t = 2.02, p < .05) revealing a significant positive influence Participants were also randomly assigned to a restorative (picture of of restorative brand WOM on self-perceptions when self-esteem is at Brawny paper towels with packaging promoting relationship with or below 4.73 (b = .29, t = 1.98, p = .05). the Wounded Warrior Project) or non-restorative (picture of Brawny PROCESS model 7 (5,000 bootstrap resamples) supported paper towels with original packaging) brand condition. All partici- moderated mediation (-.05, 95% CI = [-.1111, -.0045]). The interac- pants spread WOM by commenting on a mock Brawny Facebook tion significantly affects psychological discomfort (b = .24, t = 2.21, page. Participants rated self-perceptions of ethics using items adapt- p < .05), reducing self-perceptions (b = -.21, t = -4.47, p < .001). For ed from Reed, Aquino, and Levy (2007; α = .60). low self-esteem consumers, spreading WOM about the restorative A two-way ANOVA with threat and brand (IVs) and self-per- brand positively affects self-perceptions by reducing psychological ceptions of ethics (DV) revealed a direct effect of threat (F(1, 124) = discomfort (indirect effect: .08, 95% CI = [.0023, .2051]), supporting 13.82, p < .001; no threat: M = 5.56, SD = .81; self-threat: M = 4.97, H3 and clarifying H2. SD = 1.07), qualified by a significant threat by brand interaction (F(1, In closing, the current research reveals that following self-threat, 124) = 3.77, p = .05). Probing the interaction reveals that self-threat- spreading WOM about restorative brands can repair consumers’ self- ened consumers exhibit higher self-perceptions on the threatened at- concepts. This effect is shown to be mediated by psychological distribute after spreading WOM about a restorative (M = 5.21, SD = .96) comfort and moderated by self-esteem. It would be beneficial for fu- versus non-restorative (M = 4.67, SD = 1.14) brand (F(1, 59) = 4.01, ture research to examine how these findings extend to the purchase of p = .05). Self-perceptions of self-threatened consumers who spread restorative brands. Similarly, it would be advantageous for research WOM about a restorative brand did not significantly differ from con- to examine additional moderators to these effects and to compare the sumers who did not experience self-threat (p = .18), supporting H1. current results to instances where WOM is not spread. Study 2: Participants (n = 140, average age 36, 48% male) completed a fake IQ assessment and were randomly assigned to a REFERENCES self-threat condition. Half were told their IQ was above average (no Ariely, Dan and Michael I. Norton (2009), “Conceptual threat), half were told their IQ was below average (threat). Partici- Consumption.,” Annual Review of Psychology, 60, 475-499. pants were randomly assigned to a restorative (Special K cereal) or Belk, Russell W. (1988), “Possessions and the Extended Self,” non-restorative (Lucky Charms cereal) brand condition. All partici- Journal of Consumer Research, 15(2), 139-168. pants spread WOM by posting comments to a mock Facebook page Advances in Consumer Research (Volume 45) / 859

Berger, Jonah (2014), “Word of Mouth and Interpersonal Packard, Grant M. and David B. Wooten (2013), “Compensatory Communication: A Review and Directions for Future Knowledge Signaling in Consumer Word-of-Mouth,” Journal Research,” Journal of Consumer Psychology, 24(4), 586-607. of Consumer Psychology, 23(4), 434-450. Chung, Cindy M. Y. and Peter R. Darke, (2006), “The Consumer Reed, Americus, Karl Aquino, and Eric Levy (2007), “Moral as Advocate: Self-Relevance, Culture, and Word-of-Mouth,” Identity and Judgments of Charitable Behaviors,” Journal of Marketing Letters, 17(4), 269-279. Marketing, 71(1), 178-193. Dommer, Sara Loughran, Vanitha Swaminathan, and Rohini Rick, Scott, and Maurice E. Schweitzer (2012), “The Imbibing Ahluwalia (2013), “Using Differentiated Brands to Deflect Idiot Bias: Consuming Alcohol Can Be Hazardous to Your Exclusion and Protect Inclusion: The Moderating Role of Self- (Perceived) Intelligence,” Journal of Consumer Psychology, Esteem on Attachment to Differentiated Brands,”Journal of 23(2), 212-219. Consumer Research, 40(4), 657-675. Rosenberg, Morris (1965), Society and the Adolescent Self-Image, Elliot, Andrew J., and Patricia G. Devine (1994) “On the Princeton, NJ: Princeton University Press. Motivational Nature of Cognitive Dissonance: Dissonance as Sherer, Mark, James E. Maddux, Blaise Mercandante, Steven Psychological Discomfort,” Journal of Personality and Social Prentice-Dunn, Beth Jacobs, and Ronald W. Rogers (1982), Psychology, 67(3), 382. “The Self-Efficacy Scale: Construction and Validation,” Festinger, Leon (1957), A Theory of Cognitive Dissonance, Psychological Reports, 51, 663-671. Stanford University Press. Wicklund, Robert A. and Peter M. Gollwitzer (1981), “Symbolic Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, Self-Completion, Attempted Influence, and Self-Deprecation,” and Conditional Process Analysis: A Regression-Based Basic and Applied Social Psychology, 2(2), 89-114. Approach, New York: Guilford Press. Mandel, Naomi, Derek D. Rucker, Jonathan Levav, and Adam D. Galinsky (2017), “The Compensatory Consumer Behavior Model: How Self-Discrepancies Drive Consumer Behavior,” Journal of Consumer Psychology, 27(1), 133-146. Fight Fire with Fire: Using One Consumer Stereotype to Overcome Another via Contingency Information Julian Saint Clair, Loyola Marymount University, USA Mitchell Hamilton, Loyola Marymount University, USA Delancy Bennett, Clemson University, USA

EXTENDED ABSTRACT contingency cue. Further, we argue that the process is reliant on the Although progress has been made toward eliminating inequi- stereotypically positive link between the group and the concept to ties caused by prejudice and discrimination, underrepresentation is create the consistent, balanced cognitions necessary for conceptual- still an important issue in many fields such as Science, Technology, contingency learning; simply linking any second concept to the do- Engineering, Math, and Business (Hoyt & Murphy, 2016; Nosek, et main (e.g., science is fun) is insufficient. As such, one stereotype al., 2009; Ong, Wright, Espinosa, & Orfield, 2011). While market- (i.e., a group-attribute association) is being utilized to overcome an- ing research supports the intuition that campaigns targeting under- other. The ability of consumers to integrate cognitively inconsistent represented groups have positive effects in general (e.g., a diet soda information in a single exposure via conceptual-contingency learn- advertisement featuring an ethnic minority as a spokesperson; Aaker, ing is inexplicable using current models of associative or categorical Brumbaugh, & Grier, 2000; Luna, Ringberg, & Peracchio, 2008; learning. Stayman & Deshpande, 1989), scant research exists that explores In sum, the present work addresses the research gap surrounding diversity marketing in stereotyped domains specifically, where the counter-stereotypical marketing. We contribute to theory by deriving marketing campaigns are by definition counter-stereotypical (e.g., an predicted positive and negative effects from role models and stereo- ad for a STEM program with a female spokesperson). The present re- type threat literatures respectively, and by reconciling the competing search addresses this gap, regarding counter-stereotypical marketing, predictions with a novel theoretical model of conceptual-contingency by deriving competing predictions and building a novel theoretical learning. The model uniquely identifies conceptual-contingency cues model to reconcile them. as a moderator of whether counter-stereotypical information helps or Almost two decades of research on exposure to role models (see hurts. The integration of these alternative theoretical stances has im- Stout, Dasgupta, Hunsinger, & McManus, 2011 for a review) sug- portant implications not only for theory on stereotypes and learning, gests that counter-stereotypical advertisements may help by weaken- but also for managers and policy makers concerned with diversity. ing implicit stereotypes (Dasgupta & Asgari, 2004) or by inoculat- ing consumers against the stereotype via a strengthened self-concept Empirical Evidence (Dasgupta, 2011; Hoyt & Simon, 2011; Lockwood & Kunda, 1997, Three studies across multiple stereotypes provide converging 1999; Marx & Roman, 2002), both of which suggest a positive effect evidence for the impact of conceptual-contingency learning on the of counter-stereotypical marketing. Over two decades of research on processing of counter-stereotypical information. All studies recruit stereotype threat (for reviews see Pennington, Heim, Levy, & Larkin, participants from Amazon MTurk in exchange for compensation 2016; and Spencer, Logel, & Davies, 2016), however, suggests an (Buhrmester, Kwang, & Gosling, 2011). Study 1 (n = 125) shows alternate view: counter-stereotypical advertisements may act as cues that a stereotype-inconsistent (vs. -neutral) ad for an engineering that activate stereotype-consistent cognitions (Bargh, Chen, & Bur- program containing a quote from a female (vs. unidentified) spokes- rows, 1996; Jamieson & Harkins, 2012; Lee, Kim, & Vohs, 2011; person has a positive (null) effect on evaluations when the quote J. R. Steele & Ambady, 2006), leading to stereotype-consistent be- contains a conceptual-contingency (neutral) cue: “I found the beauty havior and thus suggesting a negative effect of counter-stereotypical [fun] in engineering;” F(1, 121) = 4.27, p = .041. This supports the marketing. contention that a counter-stereotypical ad only has a positive effect Drawing on research from information processing including in the presence of a cue that positively links the group to the domain. cognitive consistency and categorical learning (e.g., Greenwald, In the context of the ethnicity-academics stereotype (C. M. Steele, et al., 2002; Nosofsky, Palmeri, & Mckinley, 1994), we build a 1997), study 2 (n = 121) extends these findings by showing that the theoretical model that accounts for both positive and negative ef- effect reverses as the link between the conceptual-contingency cue fects of counter-stereotypical marketing via conceptual-contingency (street smarts) and the domain (higher education) is perceived to be learning. Consumers may be unable to successfully process coun- more negative, as captured by an individual difference measure; B = ter-stereotypical information because it is inherently cognitively -.526; t(121) = -3.36, p = .001. In the context of the gender-business inconsistent with the consumer’s prevailing stereotype knowledge stereotype (Powell & Butterfield, 1994), study 3 (n = 138) directly (Greenwald, et al., 2002; Heider, 1946). Thus, seeing counter-ste- manipulates the positivity of the cue-domain link with pre-tested os- reotypical information would only activate the prevailing knowledge tensible “research articles” stating that social skills may have posi- structure (e.g., the stereotype; Allen, Sherman, Conrey, & Stroessner, tive (vs. negative) effects in business. Results show that both positive 2009; Cunha, Janiszewski, & Laran, 2008), leading to stereotype- and negative effects of the cue-domain link occur, but only when consistent behavior. However, stereotyped social groups are also ste- the cue is also positively linked to the group (social skills-women); reotypically linked to a number of other concepts (e.g., “handsome F(1, 134) = 5.09, p = .026. This provides further evidence for the = male” and “beauty = female”). If consumers are able to success- proposed cognitive-consistency/balance process. fully link such a second-order concept to the domain (e.g., science Taken together, the findings show that counter-stereotypical ads is beautiful), this may allow successful incorporation of the counter- can lead to positive evaluations when they contain a conceptual-con- stereotypical information by creating a contingency (i.e., exception) tingency cue that positively links the stereotyped group to the domain rule in the knowledge structure (Nosofsky, et al., 1994). Specifically, (e.g., beauty in engineering). Lacking such a cue, results show that the new knowledge structure may be such that the stereotyped group counter-stereotypical ads may lead to negative evaluations. These re- is generally negatively associated with the domain, except in the sults demonstrate the importance of conceptual-contingency learning presence of the second-order concept that serves as the conceptual- in the processing of stereotype-inconsistent stimuli, supporting our

Advances in Consumer Research 860 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 861 theoretical model and providing important implications for research- Lockwood, P., & Kunda, Z. (1999). Increasing the salience of one’s ers, managers, and policy makers concerned with diversity. best selves can undermine inspiration by outstanding role models. Journal of Personality and Social Psychology, 76, REFERENCES 214-228. Aaker, J. L., Brumbaugh, A. M., & Grier, S. A. (2000). Nontarget Luna, D., Ringberg, T., & Peracchio, L. A. (2008). One individual, markets and viewer distinctiveness: The impact of target two identities: Frame switching among biculturals. Journal of marketing on advertising attitudes. Journal of Consumer Consumer Research, 35, 279-293. Psychology, 9, 127-140. Marx, D. M., & Roman, J. S. (2002). Female role models: Allen, T. J., Sherman, J. W., Conrey, F. R., & Stroessner, S. J. Protecting women’s math test performance. Personality and (2009). Stereotype strength and attentional bias: Preference Social Psychology Bulletin, 28, 1183-1193. for confirming versus disconfirming information depends Nosek, B. A., Smyth, F. L., Sriram, N., Lindner, N. M., Devos, T., on processing capacity. Journal of Experimental Social Ayala, A., Bar-Anan, Y., Bergh, R., Cai, H. J., Gonsalkorale, Psychology, 45, 1081-1087. K., Kesebir, S., Maliszewski, N., Neto, F., Olli, E., Park, Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of J., Schnabel, K., Shiomura, K., Tulbure, B. T., Wiers, R. social behavior: Direct effects of trait construct and stereotype W., Somogyi, M., Akrami, N., Ekehammar, B., Vianello, activation on action. Journal of Personality and Social M., Banaji, M. R., & Greenwald, A. G. (2009). National Psychology, 71, 230-244. differences in gender-science stereotypes predict national sex Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s differences in science and math achievement.Proceedings Mechanical Turk: A New Source of Inexpensive, Yet High- of the National Academy of Sciences of the United States of Quality, Data? Perspectives on Psychological Science, 6, 3-5. America, 106, 10593-10597. Cunha, M., Janiszewski, C., & Laran, J. (2008). Protection of Prior Nosofsky, R. M., Palmeri, T. J., & Mckinley, S. C. (1994). Learning in Complex Consumer Learning Environments. Rule-Plus-Exception Model of Classification Learning. Journal of Consumer Research, 34, 850-864. Psychological Review, 101, 53-79. Dasgupta, N. (2011). Ingroup Experts and Peers as Social Vaccines Ong, M., Wright, C., Espinosa, L. L., & Orfield, G. (2011). Inside Who Inoculate the Self-Concept: The Stereotype Inoculation the Double Bind: A Synthesis of Empirical Research on Model. Psychological Inquiry, 22, 231-246. Undergraduate and Graduate Women of Color in Science, Dasgupta, N., & Asgari, S. (2004). Seeing is believing: Exposure Technology, Engineering, and Mathematics. Harvard to counterstereotypic women leaders and its effect on the Educational Review, 81, 172-208. malleability of automatic gender stereotyping. Journal of Pennington, C. R., Heim, D., Levy, A. R., & Larkin, D. T. (2016). Experimental Social Psychology, 40, 642-658. Twenty Years of Stereotype Threat Research: A Review of Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Psychological Mediators. Plos One, 11. Nosek, B. A., & Mellott, D. S. (2002). A unified theory of Powell, G. N., & Butterfield, D. A. (1994). Investigating the implicit attitudes, stereotypes, self-esteem, and self-concept. Glass Ceiling Phenomenon - an Empirical-Study of Actual Psychological Review, 109, 3-25. Promotions to Top Management. Academy of Management Heider, F. (1946). Attitudes and Cognitive Organization. Journal of Journal, 37, 68-86. Psychology, 21, 107-112. Spencer, S. J., Logel, C., & Davies, P. G. (2016). Stereotype Threat. Hoyt, C. L., & Murphy, S. E. (2016). Managing to clear the Annual Review of Psychology, Vol 67, 67, 415-437. air: Stereotype threat, women, and leadership. Leadership Stayman, D. M., & Deshpande, R. (1989). Situational Ethnicity Quarterly, 27, 387-399. and Consumer-Behavior. Journal of Consumer Research, 16, Hoyt, C. L., & Simon, S. (2011). Female Leaders: Injurious or 361-371. Inspiring Role Models for Women? Psychology of Women Steele, C. M. (1997). A threat in the air - How stereotypes shape Quarterly, 35, 143-157. intellectual identity and performance. American Psychologist, Jamieson, J. P., & Harkins, S. G. (2012). Distinguishing between 52, 613-629. the effects of stereotype priming and stereotype threat on math Steele, J. R., & Ambady, N. (2006). “Math is Hard”! The effect of performance. Group Processes & Intergroup Relations, 15, gender priming on women’s attitudes. Journal of Experimental 291-304. Social Psychology, 42, 428-436. Lee, K., Kim, H., & Vohs, K. D. (2011). Stereotype Threat in the Stout, J. G., Dasgupta, N., Hunsinger, M., & McManus, M. A. Marketplace: Consumer Anxiety and Purchase Intentions. (2011). STEMing the Tide: Using Ingroup Experts to Inoculate Journal of Consumer Research, 38, 343-357. Women’s Self-Concept in Science, Technology, Engineering, Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting and Mathematics (STEM). Journal of Personality and Social the impact of role models on the self. Journal of Personality Psychology, 100, 255-270. and Social Psychology, 73, 91-103. Understanding Chat Perceptions in a Customer Assistance Channel Marion Sanglé-Ferrière, ESCP Europe, France Benjamin Voyer, ESCP Europe, UK

EXTENDED ABSTRACT been studied in great detail in the field of HCI (Yee, Bailenson, and The development of self-service technologies, in an attempt to Rickertsen 2007). make customers autonomous, have paradoxically created new situ- Customer’s decision to use chat in an assistance context can ations in which individuals may need help, for instance in the case be framed using HCI. First, Davis’ (1989) Technology Acceptance of technological failure or customer mistakes (Bitner, Ostrom, and Model (TAM) provides an effective way to predict whether a user Meuter 2002; Meuter et al. 2000). A traditional response from com- will be likely to use a given technological tool. His model puts for- panies has been to offer assistance via chat on websites, as a mean to ward two factors that predict this intention: the first one being the assist and humanize websites. Thanks to its dialogical form, chat has ‘‘perceived usefulness’’ and the second one being the “perceived ease also been presented as a way to alleviate the higher level of perceived of use” of the tool. This model has been largely applied to analyse risk encountered within e-service (Kumar and Benbasat 2002; Xu customer’s adoption of new technological tools, and has been refined 2016). over the last decades. For instance, e-service adoption studies have The concept of customer assistance has not been clearly defined added the notion of perceived risk to the initial dimensions (Feather- in the literature. We define it as the encounter between a customer man and Pavlou 2003), leaning on categories of risk identified by acknowledging a problematic situation with a product or service, Cunningham (1967) such as performance risk and psychological and a proposal for assistance by the company. Studies investigating risk. Assistance-seeking situations typically entail a certain degree chat as an assistance channel have relied on two main assumptions. of uncertainty and various degrees of stake, which is representative First, the fact that e-retail and e-service need to be humanized (e.g. of risky situations. In their study on channel choice in customer-ini- Hassanein and Head 2007; Park, Chung, and Rutherford 2011). Sec- tiated communication, Polo and Sese (2016, 278) argue that amount ond, the fact that it is perceived as more risky by consumers than of perceived risk, which differs depending on the nature (channel) of face-to-face interactions (Paluch and Wünderlich 2016; Feather- the interaction, will “determine customer channel preferences”. Per- man and Pavlou 2003). Moreover the limited studies looking at chat ceived risk is thus also likely to play a role in a customer assistance have taken a company’s point of view, rather than a consumer one. channel decision. As a result, the research field lacks an understanding of customers’ perceptions of chat. This research addresses this gap and aims at i) Methodology: Semi Structured Interviews Investigating understanding perceptions of chat as an assistance channel, and ii) “Life Stories” explore customers’ motivations and limiting factors for the use of This research adopts a qualitative approach to explore and un- chat in an assistance context. Doing so, it addresses recent calls for derstand customers’ perceptions. We conducted 23 semi-structured research on customer experience in context (Klaus 2013) and on interviews (totalling to 1165 minutes), using a purposive sampling customer-initiated communications, which go beyond a purchasing strategy. Diversity was sought in terms of age, gender and type of context (Polo and Sese 2016). occupation, in order to obtain the most diverse customer experienc- This research investigates customers’ perceptions of online chat es. 17 interviewees recalled having used chat for online assistance. as an assistance channel in their own real-life assistance context. We These were asked to recall one (or more) chat experience with a adopt a service and consumer-centred point of view to study chat and customer service representative online – whether live chat or vir- customer’s need for assistance, relying on Human Computer Inter- tual agent. They were encouraged to generate “life stories”, based action (HCI) theories. We adopt a qualitative research design using on a narrative introspection (Carù, Cova, and Pace 2014). Six did semi-structured interviews. Our findings identify key criteria, which not recall ever using chat as an online assistance channel. For these, determine the need for assistance and willingness to use chat as per- interviews focused on their experiences of customer assistance (e.g. ceived by customers. trying to solve a service failure). In addition, they were called out to imagine how assistance by chat might have been helpful – or not – in Literature Review: Live Chat vs . Virtual agents, these situations. The use of a projective technique facilitate custom- Technology Acceptance, and Perceived Risks ers’ verbalisation of the obstacles that he / she imagines concerning Chat has been defined in the service literature as the “sending the use of chat (Lavorata, Nilles, and Pontier 2005). Interviews, tran- and receiving of short text-based messages where the sender and the scripts, and analyses were conducted in French, and subsequently recipient communicate usually with no (or minimal) delays” (Froeh- translated in English by the first author. Interviews were coded sepa- le 2006). A main limitation of this definition is that it does not con- rately using thematic analysis, and categories emerged thanks to consider the nature of the interlocutor – whether the interlocutor is a hu- stant comparison over the data (Corbin and Strauss 2015). man being or a virtual agent. Customers, however, may not be aware, before starting a conversation, whether their interlocutor will be of Findings: Heterogeneity of Customer’s Perceptions and a human or virtual nature. Chat, in a customer assistance context, Appreciations of Chat usage in Assistance Situations generally comes in two distinct forms. First, live chat refers to chat Respondents point out the perceived nature of the interlocutor between humans. Live chat is often assimilated to a generic form of as key in their perceptions and interpretations. They characterize chat computer-mediated communication, and scholars draw from theories features according to two dimensions: a performance one and a rela- such as media richness theory (e.g. Froehle 2006; Graetz et al. 1998) tional one. We also identified four main types of motivations, which or social presence theory (e.g. Verhagen et al. 2014) to study it. Live guide respondents through their assistance experience. The impor- chat is considered synchronous text and a rather low media in terms tance of the nature of the chat interlocutor of its richness (Froehle 2006). Second, virtual agents, which are artificial-intelligence-operated chat-robots (or chatbots), and have

The question of the nature of the interlocutor lies at the heart influenced (e.g. feeling obliged to buy a product or service), and iv) of the perception of chat in an assistance context. As summarised by to feel in control over the course of the interaction (e.g. being able to one of the respondent: interrupt it easily). Thus, chat alleviates risks traditionally associated with interacting with a representative. It is somehow mysterious, I am not sure whether it is like Siri For respondents who dealt with virtual agents, the evaluation of in a written manner or if real people write themselves. When I relational features of chat were overall negative. In this case, partici- see these chat proposals, I wonder: “Is there really someone pants perceive virtual agents to lack the capacity to understand cus- behind?” Maybe it depends on websites. (Mathilde, 25) tomers. This was often interpreted as a lack of consideration from the company. On some occasions, when participants reported a sudden Respondents often express doubts about the true nature of their realisation that they had been conversing with a virtual agent rather interlocutor – i.e. human vs. computer – when discussing a past chat than a human representative, they generally felt deceived. experience. When the interlocutor is perceived as being a virtual agent, participants elaborate on the fact that companies may not be Chat as a customer assistance channel: customers’ willing to use humans either because it would be too costly or be- motivations cause companies are not genuinely willing to help customers. Chat As for the motivations that led customers to choose a specific interlocutors identified as virtual agents generated mostly negative assistance channel, four main categories emerged: assistance request appreciations. When chat interlocutor is identified as a human, par- avoidance, impression of having no choice, performance seeking and ticipants describe the interaction as “flowing freely”, they feel con- minimum direct contact. The assistance-request avoidance customer sidered as dealing with a dedicated interlocutor. is deferring as much as possible the moment where they will look for assistance, trying to solve problems autonomously first. The other The perceived performance of customer assistance via chat three categories gather customers, who more readily accept to look Interviewees perceived the overall performance of chat in terms for assistance. Some customers report an impression of having no of two main dimensions: i) the outcome of the conversation, and ii) choice. For these, chat might be the only option offered (e.g. those on the time required to find a solution. a low cost mobile phone contract) or they might have become used to When reflecting on the outcome of a chat experience, most cus- using a specific channel (phone or chat) as a default option to contact tomers assessed the performance and usefulness of chat based on companies. In sum, some respondents do not feel that they are mak- their perceptions of the nature of their interlocutor and of the compa- ing a conscious channel choice. The performance-seeking customer ny’s perceived intention to help or not. A chat interlocutor perceived is consciously setting up a performance-oriented strategy, based on as a virtual agent, does not, typically, seem useful to some: former positive experiences, what they know about the company and I feel it [chat] is not going to be of any use, I still believe chat what they anticipate about the complexity, importance and emer- is a machine. If a machine with the picture of a woman asks gency of the situation. Finally, the minimum direct contact seeking me « What are you looking for? » and that you get ready made customers, though willing to be assisted, drive their contact strategy answers like FAQs… I would rather look by myself. (Nelly, 40) towards minimizing direct interactions, sometimes at the expense of performance. Respondents also reported a perceived need to have some spe- Overall, some customers feel genuinely motivated by the avoid- cific abilities or skills – for instance being able to use specific terms ance of direct interaction with the company. Chat seems to somehow - when chatting with a virtual agent. Conversely, chatting with a address this concern by offering a more distant and customer-con- human is generally perceived as easier. Some respondents reported trolled form of communication. However chat also raises heteroge- communication failures, mostly when interacting with virtual agents, neous evaluations from customers, these differing mostly based on when they did not succeed to receive valuable information. their interlocutors’ perceptions (human vs. virtual agent, see above). In terms of the time dimension, overall, respondents put forward a series of performance characteristics of chat, in comparison Discussion, Limitations, Future research and Conclusion to other assistance channels: its free nature, its immediacy, ease of Discussion access and initiation – especially the fact that, compared with phone The contribution of this article to this emerging field of research customer assistance, a representative can be reached directly, with- is threefold. First, and paradoxically, results suggest that some re- out having to listen to an option menu first – and its flexibility allow- spondents put forward the lack of richness of chat as one of its main ing multitasking. assets. Two properties of chat appear beneficial to customers: i) the The relational dimension of chat fact that chatting communicates fewer cues in the conversation, Chat is predominantly seen as less spontaneous and less warm which lessens the risk of negative emotions. This can be seen as an than a conversation on the phone or face-to-face, as it conveys fewer illustration of the ‘neutrality effect’ (Byron 2008), which refers to emotional cues. Some respondents paradoxically perceived this dis- the fact that a media providing less non-verbal cues generally cre- tance as an asset in an assistance context: ates less arousal. This neutrality effect, seen as positive feature in the context of assistance, appears to challenge the assumption that Chat is perfect, you have all advantages of email, like being humanisation and social richness should always be sought for within more anonymous than phone, not to have to discuss with some- customer-company relationships. And ii), the fact that chat is a less one, but it is more immediate and you can react to it live (Flo- formalized way of communicating, which gives back some control rent, 28). and freedom to customers. This, in turn, contributes to reducing threats associated with assistance situations, such as loosing face. More specifically, for some customers, the lack of richness of In sum, customers seeking assistance via chat may do so because of chat as an assistance channel means that customers tend i) not to the fact that chat offers a way to focus on a specific problem, without feel judged by a perceived condescending tone of voice, ii) not to be caring about the form or richness of the interaction. directly exposed to a representative’s disengagement, iii) not to feel 864 / Understanding Chat Perceptions in a Customer Assistance Channel

Second, chat can be perceived as a way to lessen some of the Featherman, Mauricio S., and Paul A. Pavlou (2003), “Predicting performance and psychological risks associated with assistance E-Services Adoption: A Perceived Risk Facets Perspective.” seeking. In terms of performance risk, chat is perceived as a way International Journal of Human-Computer Studies 59 (4): to reduce time spent on assistance. Following Bellman, Lohse, and 451–74. Johnson (1999) time consideration is an important driver for e- Froehle, Craig M. (2006), “Service Personnel, Technology, and services usage. In terms of psychological risks, the neutrality and Their Interaction in Influencing Customer Satisfaction.” specific form of interaction of chat alleviates the risk of losing face Decision Sciences 37 (1): 5–38. in an interaction. One of the contributions of this research is thus to Graetz, Kenneth A., Edward S. Boyle, Charles E. Kimble, Pamela better understand consumers who are seeking to avoid contacts in an Thompson, and Julie L. Garloch (1998), “Information Sharing assistance situation. This statement challenges the idea that e-service in Face-to-Face, Teleconferencing, and Electronic Chat is necessarily perceived as riskier than services implying direct in- Groups.” Small Group Research 29 (6): 714–43. teraction. Hassanein, Khaled, and Milena Head (2007), “Manipulating Third, this study shows that the nature of the chat interlocutor in Perceived Social Presence through the Web Interface and Its an assistance context can be elusive albeit of the greatest importance Impact on Attitude towards Online Shopping.” International for customers. The perceived nature (human or artificial intelligence) Journal of Human-Computer Studies 65 (8): 689–708. is critical to understand the adoption of this technological media. Klaus, Philipp (2013), “The Case of Amazon.com: Towards Thus, this study contributes to the technology acceptance literature a Conceptual Framework of Online Customer Service by confirming perceived risk as a factor influencing channel/technol- Experience (OCSE) Using the Emerging Consensus Technique ogy choice (Featherman and Pavlou 2003; Polo and Sese 2016). (ECT).” Journal of Services Marketing 27 (6): 443–57. Kumar, Nanda, and Izak Benbasat (2002), “Para-Social Presence Limitations and future research and Communication Capabilities of a Web Site: A Theoretical This study carries two main limitations. First, experiences and Perspective.” E-Service Journal 1 (3): 5–24. representations of chat are likely to evolve rapidly, as consumers get Lavorata, Laure, Jean-Jacques Nilles, and Suzanne Pontier (2005), additional opportunities to experience chat in a customer assistance “La Méthode Des Scénarios: Une Méthode Qualitative context. Second, the artificial intelligence used by virtual agents is Innovante Pour Le Marketing. Application Au Comportement likely to improve dramatically in the coming years, which could Éthique Du Vendeur En B to B.” Décisions Marketing, no. 37: eventually also change perceptions and representations. 67–75. Future research should focus on a wider non-chatters popula- Meuter, Matthew L., Amy L. Ostrom, Robert I. Roundtree, tion to better understand the barriers to the use of chat. Moreover, it and Mary Jo Bitner (2000), “Self Service Technologies: would be interesting to investigate chat perceptions in specific con- Understanding Customer Satisfaction with Technology-Based texts (e.g. pre-purchase, after sale, technological failure, etc.) as they Service Encounters.” Journal of Marketing 64 (July): 50–64. may vary accordingly. Finally, the concept of “need for customer Paluch, Stefanie, and Nancy V. Wünderlich (2016), “Contrasting assistance” appears pivotal to understand assistance seeking behav- Risk Perceptions of Technology-Based Service Innovations in iours, and could be further explored and operationalized in order to Inter-Organizational Settings.” Journal of Business Research evaluate customers’ propensity to seek assistance, and engage with 69 (7): 2424–31. chat. Park, JungKun, HoEun Chung, and Brian Rutherford (2011), REFERENCES “Social Perspectives of E-Contact Center for Loyalty Building.” Journal of Business Research 64 (1): 34–38. Bellman, Steven, Gerald L. Lohse, and Eric J. Johnson (1999), Polo, Yolanda, and F. Javier Sese (2016), “Does the Nature of the “Predictors of Online Buying Behavior.” Communications of Interaction Matter? Understanding Customer Channel Choice the ACM 42 (12): 32–38. for Purchases and Communications.” Journal of Service Bitner, Mary Jo, Amy L. Ostrom, and Matthew L. Meuter (2002), Research 19 (3): 276–90. “Implementing Successful Self-Service Technologies.” The Verhagen, Tibert, Jaap van Nes, Frans Feldberg, and Willemijn Academy of Management Executive 16 (4): 96–108. van Dolen (2014), “Virtual Customer Service Agents: Using Byron, Kristin (2008), “Carrying Too Heavy a Load? The Social Presence and Personalization to Shape Online Service Communication and Miscommunication of Emotion by Encounters.” Journal of Computer-Mediated Communication Email.” Academy of Management Review 33 (2): 309–27. 19 (3): 529–45. Carù, Antonella, Bernard Cova, and Stefano Pace (2014), Xu, Jingjun (David) (2016), “Retaining Customers by Utilizing “Combining Qualitative Methods in Practice: A Technology-Facilitated Chat: Mitigating Website Anxiety Contextualized Account of the Evolution of Consumer and Task Complexity.” Information and Management 53 (5): Studies.” Edited by Professor Simone Guercini. Management 554–69. Decision 52 (4): 777–93. Yee, Nick, Jeremy N. Bailenson, and Kathryn Rickertsen (2007), Corbin, Juliet M., and Anselm L. Strauss (2015), Basics of “A Meta-Analysis of the Impact of the Inclusion and Realism Qualitative Research: Techniques and Procedures for of Human-like Faces on User Experiences in Interfaces.” In Developing Grounded Theory. 4. ed. Los Angeles: SAGE. Proceedings of the SIGCHI Conference on Human Factors in Cunningham, Scott M. (1967), “The Major Dimensions of Computing Systems, 1–10. ACM. http://dl.acm.org/citation. Perceived Risk.” Risk Taking and Information Handling in cfm?id=1240626. Consumer Behavior 1: 82–111. Davis, Fred D. (1989), “Perceived Usefulness, Perceived Ease of Use, Ans User Acceptance of Information Technology.” MIS Quarterly 13 (3): 319–40. Do Financial Restrictions Deteriorate or Improve Self-Control? The Role of Mindfulness Gülen Sarial Abi, Bocconi University, Italy Sadaf Mokarram Dorri, Bocconi University, Italy

EXTENDED ABSTRACT In study 2, we replicated the findings of study 1 using again Perception that your monetary situation does not let you to an online panel. After the financial restrictions manipulation as in consume what you desire or want is something prevalent that each study 1, we asked participants to imagine themselves in two situa- individual might experience in life. This paper examines how the tions taken from Dewitte, Bruynel and Geysken (2009), where they duration of financial restrictions influence self-control behavior of need to decide between ordering a healthy versus unhealthy option individuals. in order to measure their self-control performance. In this study, we One stream of research on financial constraints demonstrate demonstrated that those who have short-term financial restrictions different coping strategies that people might use such as consider- also behave impulsively in an unrelated domain compared to those ing opportunity costs (Spiller 2011), efficiency or priority planning who have long-term financial restrictions (t(143) = 2.105, p = .037). (Fernbach et al. 2015) or preferring material goods over experiences In study 3, we tested our predictions again in an unrelated do- (Tully et al. 2015). All these findings suggest that those who are fi- main by comparing preference for healthy versus unhealthy products nancially constrained are somewhat mindful (i.e., being aware of in eating domain for both short-term and long-term financial restric- the present-moment and accept it non-judgmentally) of their current tion conditions. The result showed that those who have financial re- situation which helps them in controlling themselves from engag- strictions for a shorter period of time prefer more unhealthy food ing in activities that might have positive consequences now but do over healthy food compared to those who have financial restrictions not lead to long-term benefits. However, another stream of research for a longer period of time (t(109) = -1.668, p = .098). Furthermore, suggest the negative effect of financial restrictions on self-control we asked participants to complete the Mindful Attention Awareness performance, as those who have scarce resources (e.g., financial re- Scale (Brown and Ryan 2003). We have also demonstrated mind- sources) focus single-mindedly on managing the scarcity at hand and fulness as the underlying mechanism explaining the effect of stage they might neglect other and even more important things (Mullaina- of financial restriction on self-control (ab=-.1855, 90% confidence than and Shafir 2013). interval [CI] = [–.43, –.04]). In this paper, we provide a solution to these conflicting results In study 4, we demonstrated the effects in a behavioural con- about the effects of being financially restricted on self-control. We text. Upon arrival to the lab, participants were randomly assigned to differentiate between duration of financial restrictions and suggest the mindfulness or control conditions; we manipulated mindfulness that these different findings actually refer to either having short-term by asking participants to meditate following the instructions of the or long-term financial restrictions. More specifically, we suggest that mindfulness app for fifteen minutes on a yoga mat. Participants in when someone perceives him/herself to be financially restricted for a both meditation and control conditions were assigned randomly to short period of time, the person would not be mindful about the situa- short-term or long-term financial restrictions conditions. After read- tion, which would deteriorate self-control. However, we suggest that ing two scenarios adapted from Spiller (2011), participants engaged the results would not be the same for those who perceive financial in a writing task, to reinforce the duration of the financial restriction restrictions for a longer period of time. Imagine you go out for shop- (i.e., short-term versus long-term) manipulation. Next, participants ping. You like several items, having in mind that you have financial were shown the menu of the bar and they were asked to choose the restrictions (i.e., you have mortgage bills to make, your expenses items that they would purchase. For each item, participants were for the month are over your budget, etc.). Hence, you are mindful of provided with calories and cost information. In this study, we dem- your current financial situation. You control yourself and go to the onstrated that while in the control condition, we replicated our key cashier with the items that you would like to purchase (self-control findings so that those who have financial restrictions for a shorter pe- task 1). You make the payment and then the cashier suggests you an riod of time engage in behaviour that is low in self-control compared offer with a discount on an item (self-control task 2). In this case, be- to those who have financial restrictions for a longer period of time (F cause you are mindful of your current financial situation, you would (1, 93) = 20.94, p < .001), the results were not the same for those who simultaneously experience the second self-control task with the first- were in the mindfulness condition (p = .218). In the mindfulness con- one and you would be better off resisting the offer. Hence, we suggest dition, the effect attenuated so that self-control behaviour did not sig- that you would be better in self-control. nificantly differ among the different stages of financial restrictions. We first preliminary tested our prediction using archival data of The results of this paper have implications for self-control, fi- Bank of Italy for a period between 2006 and 2014. More specifically, nancial restrictions, and mindfulness literature by demonstrating the using the archival data, we demonstrated that those who have finan- effect of stage of the financial restrictions on self-control. cial restrictions for a longer period of time (a) spend less for both durable and non-durable goods (β =-.02, p < .001) and (b) have more REFERENCES deposit savings (p = .069), controlling for the income and liabilities. Brown, Kirk Warren and Richard M. Ryan (2003), “The Benefits In study 1, using an online panel, we randomly assigned par- of Being Present: Mindfulness and Its Role in Psychological ticipants to short-term or long-term financial restrictions writing-task Well-Being,” Journal of Personality and Social Psychology, conditions. The result demonstrated that those who have financial 84(4), 822-48. restrictions only for a short-period of time are more willing to pay Dewitte, Siegfried, Sabrina Bruyneel, and Kelly Geyskens (2009), for products that are offered to them impulsively compared to those “Self-Regulating Enhances Self-Regulation in Subsequent who have financial restrictions for a longer period of time (t(130) = Consumer Decisions Involving Similar Response Conflicts,” 1.86, p = .065). Journal of Consumer Research, 36, 394-405.

Fernbach, Philip M., Christina Kan, and John G. Lynch Jr. (2015), Tully, Stephanie M., Hal E. Hershfield, and Tom Meyvis (2015), “Squeezed: Coping with Constraint through Efficiency and “Making Your Discretionary Money Last: Financial Prioritization,” Journal of Consumer Research, 41, 1204-1227. Constraints Increase Preference for Material Purchases by Mullainathan, Sendhil and Eldar Shafir (2013). Scarcity: Why Focusing Consumers on Longevity,” Journal of Consumer Having Too Little Means So Much. Time Books: Henry Holt Research, 42, 1, June, 59-73. & Company LLC. New York: NY. Spiller, Stephen A. (2011), “Opportunity Cost Consideration,” Journal of Consumer Research, 38, 595-610. Those Who Have Financial Restrictions Cannot Buy but Can Save Time: I Have to Do Everything by Myself BUT with One Product Fits it All Gülen Sarial Abi, Bocconi University, Italy Zeynep Gürhan Canli, Koc University, Turkey

EXTENDED ABSTRACT from Etkin et al. 2015). We combined these measures to form a time Although previous research investigates how having one’s perceptions index (α = .55). monetary situation as scarce influences one’s behaviour or how As predicted, a one-way analysis of variance (ANOVA) on having one’s time as scarce influences one’s behaviour, previous re- subjective time perceptions showed that having financial restrictions search is silent on how these two scarce resources might interact. In made people to feel that their time is restricted (Mfinancial = 4.82, SD = this research, we investigate how those who are financially restricted 1.39 versus Mcontrol = 4.08, SD = 1.21, t(127) = 3.22, p = .002). perceive their time and how their perception of time influences their In study 2 (N = 123), we replicate the results of study 1 and test product preferences. the effect of stress and anxiety as the mechanism that explains the How do people who have financial restrictions perceive time? effect of financial restrictions on time perceptions. First, we manipu- Previous research signals that those who have financial restrictions lated salience of financial restrictions as in study 1. Second, we mea- would perceive their time as restricted as well. On one hand, previous sured feelings of stress and anxiety as in Etkin et al. (2015). More research suggests that because those who have financial restrictions specifically, participants reported their current stress and anxiety would perceive stress (Sarial-Abi et al. 2016), they would perceive levels on seven-point scales (1 = “very little” and 7 = “a lot.”). The their time as limited, which would result as paying more for time- measures were highly correlated (r = .85) and combined. Third, we buying activities (e.g., expedited shipping; Etkin, Evangelidis, and measured subjective time perceptions. To increase the generalizabil- Aaker 2015). Another stream of research suggests that because one ity of our findings, we added more items to the questions from those focuses single-mindedly on the scarce resource (i.e., money), one in Study 1. Other than the measures that we had in study 1, we also cannot consider other important resources (Mullainathan and Shafir asked participants to indicate their agreement with four other state- 2013). As a result, those who have financial restrictions might end up ments (Etkin et al. 2015; Rudd et al. 2012): “I am in a rush,” “I don’t not efficiently managing their time, resulting in neither time-buying have enough time,” “Time is slipping away” and “I am pressed for nor time-saving activities. In this paper, we suggest that those who time” on different seven-points scales (1 = “not at all” and 7 = “very have financial restrictions perceive the time as being more restricted much”). All items were then combined to form a time perceptions than those who have no financial restrictions when they experience index (α = .76). more stress and anxiety compared to those who do not have finan- As predicted, a one-way analysis of variance (ANOVA) on time cial restrictions. When individuals perceive time as restricted, they perceptions showed that those who have financial restrictions per- might engage in activities to save time. We show that those who have ceived the time as more restricted (Mfinancial = 4.19, SD = 1.19) than financial restrictions might save time by purchasing and using multi- those in the control condition (Mcontrol = 3.82, SD = 1.13, t(121) = purpose products. 1.78, p = .078). To test the proposed underlying process, we also used In study 1 (N = 129), we examine whether having financial bias-corrected bootstrapping to generate 95% confidence interval restrictions makes consumers feel time as limited. We manipulated around the indirect effect of stress and anxiety. The analysis revealed salience of financial restrictions and measured subjective time per- a significant indirect effect (ab = -.36, 95% confidence interval [CI] ceptions. We have asked participants in the financial restrictions = [–.69, –.05]). condition to think about the factors that contribute to their financial In study 3, we tested our prediction using the results of The restrictions (Tully et al. 2015). More specifically, participants in the Survey on Household Income and Wealth (SHIW) that is conduct- financial restrictions condition read: ed by Bank of Italy since 1960. For the purposes of this study, we analysed the data starting from 2006 until the most recent one (i.e., Everyone has financial constraints in their lives, but the factors 2014). The survey reports consumption of several types of insurance that contribute to these constraints tend to vary. What are the options. More specifically, it asks respondents whether they possess factors that require you to be careful with how you spend your any of the following insurance funds: a) life products with profits, money? What limits your monthly discretionary income? In- b) unit-linked insurance funds, c) index-linked insurance funds, d) clude the aspects of your current situation that most contribute capitalization products, e) multi-branch insurance funds, f) open pen- to your financial constraints (e.g., mortgage or rent, family ex- sion fund, and g) death insurance and other pure risk. Based on the penses, uncertainty of future income, health care costs, student characteristics of each of the previous insurance funds, we classi- loans, lack of income, limited savings, bills that need to be paid, fied them into single-purpose use or multi-purpose use. -Our data expensiveness of entertainment . . . ). set includes 5,992 observations from 2006 to 2014. As a proxy for multi-purpose insurance consumption, we used a dummy variable Participants were asked to write minimum of 250 characters and which takes the value 1 if individuals have chosen a multi-branch write in detail as much as possible. insurance, general life products insurance, or an open pension fund, and zero if they have chosen a single-purpose insurance option (i.e., To provide a similarly demanding task for the participants in the unit-linked insurance, index-linked insurance, capitalization prod- control condition, we asked them to list 10 facts that they knew ucts insurance, death insurance and other pure risk insurance). We to be true (Tully et al. 2015). controlled for the individual income with the logarithm of net in- Next, we measured subjective time perceptions. Participants reported come. As a measure for financial restrictions we used the total sum of their perception that time was expanded (reverse-scored), boundless liabilities that individuals have. To test our hypothesis, we estimated (reverse-scored), constricted, and that they do not have enough time the following panel regression: MP = β + β z + β inc +β fr +ε on seven-point scales (1 = “not at all,” and 7 = “very much”; adapted 0 1 2 3 Advances in Consumer Research 867 Volume 45, ©2017 868 / Those Who Have Financial Restrictions Cannot Buy but Can Save Time: I Have to Do Everything by Myself BUT with One Product Fits it All

Where MP is the indicator of multi-purpose choice, z is a vector ship between financial restrictions and time perceptions. More spe- of socio-demographic variables, inc is the log of financial income,fr cifically, we demonstrate that those who have financial restrictions is the log of liabilities of the respondent, and ε is a standard normal perceive the time as more scarce compared to those who have no random error. financial restrictions. Second, we demonstrate that stress and anxiety We tested our prediction using a probit model utilizing data are the underlying mechanism that explains the effect of financial from 5,992 observations over an 8-years period. The probability of restrictions on time perceptions. Third, we demonstrate that those choosing multi-purpose insurance options was higher than the prob- who have financial restrictions cannot buy time when they perceive ability of choosing single-purpose insurance options, when individu- the time as scarce. Rather, they save time to cope with their scarce als have higher financial restrictions (β = .04, p = .054), controlling time perceptions. We show that they save time by preferring multi- for the net income level of individuals and their demographic char- purpose items compared to those who have no financial restrictions. acteristics. In study 4 (N = 206), we randomly assigned participants to fi- REFERENCES nancial restrictions and control conditions and tested the effects on Etkin, Jordan, Ioannis Evangelidis, and Jennifer Aaker (2015), preference of several multi-purpose items. Results supported the “Pressed For Time? Goal Conflict Shapes How Time is Perceived, Spent, and Valued,” Journal of Marketing predicted effects (for multi-purpose shampoo: Mfinancial = 4.42, SD Research, 52 (June), 394-406. = 3.11 versus Mcontrol = 3.76, SD = 2.17, t(204) = -1.67, p = .096; Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too multi-purpose printer: Mfinancial = 110.01, SD = 120.023 versus Mcontrol little means so much. Macmillan. = 79.34, SD = 64.17, t(204) = -2.10, p = .037; cake mix: Mfinancial = Sarial-Abi, Gülen, Zeynep Gürhan-Canli, Tarcan Kumkale, and 3.44, SD = 3.00 versus Mcontrol = 2.56, SD = 2.61, t(204) = -2.16, p = Yeosun Yoon (2016), “The Effect of Self-Concept Clarity on .032; acne treatment: Mfinancial = 7.02, SD = 5.95 versus Mcontrol = 5.54, Discretionary Spending Tendency,” International Journal of SD = 4.41, t(204) = -1.92, p = .057; multi-purpose cooker: Mfinancial = Research in Marketing, 33(3), 612-623. 71.69, SD = 49.21 versus Mcontrol = 57.60, SD = 67.91, t(204) = -1.72, Tully, S. M., Hershfield, H. E., & Meyvis, T. (2015). Seeking p = .086; multi-purpose detergent: Mfinancial = 7.02, SD = 5.95 versus Lasting Enjoyment with Limited Money: Financial Constraints Mcontrol = 5.54, SD = 4.41, t(204) = -1.92, p = .057). Results of this paper contribute to the literature on financial Increase Preference for Material Goods over Experiences. restrictions, time, and product preferences. First, we contribute to Journal of Consumer Research, 42(1), 59-75. the literature on financial restrictions by demonstrating the relation- Monsters in Our World: Narrative Transportation in Pokémon GO’s Mixed Reality Joachim Scholz, Cal Poly, USA Andrew Smith, Suffolk University, USA

EXTENDED ABSTRACT The emplacement of story content in consumers’ physical sur- Storytelling is a central part of the consumption experience (van roundings is at odds with core assumptions of narrative transporta- Laer et al. 2014). Traditionally, consumers are exposed to brand sto- tion theory, which posit the story world as entirely separate from the ries through advertisements (e.g, Phillips and McQuarrie 2010) and physical and social reality of everyday life (van Laer et al. 2014). social media (e.g., van Laer and de Ruyter 2010). Recent innovations Furthermore, the act of transporting is typically described as creating in augmented reality (Scholz and Smith 2016) and geo-media (La- a narrative in the story receiver’s head. While this emphasizes the penta 2011) have provided marketers with additional opportunities to agency of the story receiver, it also characterizes the task of narrative tell their brand stories: ones that are emplaced in consumers’ every- transportation as a purely cognitive and emotive process of creat- day environments. For example, transit riders might find themselves ing mental imagery associated with the story’s text. The paradox of in the presence of zombies as part of an augmented reality campaign geo-mediated storytelling is thus that the world of origin, which is to promote the TV show The Walking Dead. supposed to be left behind through a cognitive process, is exactly the While emerging spatial media offer tremendous potential for same place where the story unfolds, and where story receivers create telling brand stories, the ways in which consumers respond to these a narrative while being embodied in the mixed reality of the story- geo-mediated narratives are little understood. Previous research has world/world-of-origin hybrid. introduced narrative transportation theory (Green and Brock 2002) to Given the emplaced nature of stories told via geo-media, and explore how consumers are drawn into a narrative and enjoy a story. the embodied nature of how these stories are consumed, our research However, as we will describe below, central assumptions of narra- aims to explore how narrative transportation is experienced in mixed tive transportation theory are called into question when the story is realities. In particular, we are interested in how story elements such told through a nexus of GPS-enabled mobile devices, location-based as characters and plot lines are experienced as “real”, and more gen- content, and augmented reality technologies that create multiple re- erally how the story world becomes stabilized in the physical world alities through “the subjective coming togethers in time and space of and accepted as part of everyday life. material and virtual experiences” (Graham, Zook and Boulton 2012, p. 465). Conceptual Lens: Mixed Realities Pokémon GO is a well-known story situated in this type of Mixed realities describe hybridized experiences in which vir- mixed reality. In this research, we explore the mechanics of narrative tual content is inexorably linked with physical materialities through transportation when a story is told and experienced through geo-me- a mutual constitution of media, content and space (Farman 2012; dia that create a mixed reality. We present findings from a qualitative Graham et al. 2012). Human geographers and communications schol- study on Pokémon GO and contribute understanding about narrative ars have suggested “geo-media” (Lapenta 2011) and “spatial media” transportation in the context of this new storytelling approach. (Leszczynski 2015) as terms that describe the nexus of location- aware devices, geo-coded content, and augmented reality technolo- Narrative Transportation gies that gives rise to mixed realities, in - for example - large-scale The role of stories in consumer behavior has frequently been ex- games such as Pokémon GO and Ingress that constantly interweave plored through the concept of narrative transportation (Gerrig 1993; the story of the game with the actuality of the physical world where Green and Brock 2002), which describes the process of getting lost the game is played (Chess 2014). in a story. This detachment from reality and immersion in a narrative Mixed realities research thus recognizes the important role of world has been shown to affect consumers’ attitudes and intentions (narrative) content, for example, when exploring how “code blends (van Laer 2014), as well as their enjoyment and engagement with a content and place” (Graham et al. 2012, p. 468). In particular, previ- story (Green, Brock and Kaufman 2004). ous mixed realities research has detailed how geo-mediated content Previous research has predominantly explored narrative trans- produces unstable representations of space that are always ‘of-the- portation in the context of audio and/or visual media, while inter- moment’ and remade through the embodied, social, and technologi- active media such as computer games have received comparatively cal practices of those that produce these mixed realities. Content less attention (Green and Jenkins 2014). Some studies have explored shapes the meaning of certain places, for example, when global infra- narrative transportation in completely virtual environments such as structure is made invisible in Ingress, and patterns of social relations, computer games (Elson et al. 2014) and virtual reality (Ahn, Le and for example, when Ingress players need to interact with strangers to Bailenson 2013), but no study, to our knowledge, has thus far ex- achieve certain game-related goals (Chess 2014). amined narrative transportation in the context of mixed realities that emplaces the story in the physical environment of the story receiver. Context & Methods Geo-mediated stories are similar to stories told via digital games We investigate geo-mediated narrative transportation in the and virtual environments by virtue of their highly interactive and context of Pokémon GO. Pokémon is a narrative brand centered social nature (Biocca 2002). However, the glaring difference is that around the adventures of Ash — the Pokémon trainer — and an as- geo-mediated stories are emplaced in the physical environment of sortment of monsters. The entire Pokémon experience — collecting, the receiver. Unlike digital games or virtual environments where the training, and battling against other Pokémon — has been confined to story is told on digital screens or in self-contained heads-up displays, this fictional world, until the release of Pokémon GO in July 2016. respectively, geo-mediated stories unfold in the everyday spaces of Pokémon GO uses geo-location technology to merge consumers’ re- consumers. al-world environments (e.g., the city they are walking through) with game-relevant content such as landmarks (i.e., “Pokéstops”), activ-

Advances in Consumer Research 869 Volume 45, ©2017 870 / Monsters in Our World: Narrative Transportation in Pokémon GO’s Mixed Reality ity zones (i.e., “gyms”), and most crucially brand story characters parallels between story content and the physical space in which it (i.e., Pokémons). It allows consumers to take on the role of Ash, the is embedded, helping to bolster perceived story verisimilitude. trainer, and live out the Pokémon experience in real life. For example, they search for, and express delight about, instances We study narrative transportation in the context of Pokémon in which Pokémon characters appear to fit realistically in the spa- GO through unstructured, in-depth interviews. We interviewed 13 tial environment (e.g. when they are sitting on a sofa) (Quote 2.2). respondents in August and September, 2016, and focused our discus- Consumers even engage in more imaginative sensemaking to help sion, broadly, on understanding consumer experience with Pokémon construct the vision of a hybridized reality. For example, they draw GO. The interviews lasted, on average, 60-90 minutes, and were on their knowledge of the physical world to explain aspects of the transcribed, coded, and analyzed. It was through the iterative coding story world, such as the localized presence of Pokémon characters and analysis process that we came to focus analytically on narrative in certain places, which they attribute to the existence of nests, a transportation. To further inform our understanding of the phenom- concept not integrated into Pokémon GO. In doing so, consumers act enon and data, we also actively played the game and read media in co-creative capacity to help elaborate on the ontology of the story coverage. world, and make it seem more plausible. In both examples, the fact that the narrative content adheres to the properties of the physical Findings world helps to spatially validate the story world for consumers. Geo-media enable storytellers to emplace narratives in the physical world amidst a nexus of spatial and social relations. Under Social Story Experiences these conditions, transportation morphs from being a mental journey Geo-media embeds stories in the physical world, and that in an imaginary, external world into a joint cognitive and embod- means those stories often have bearing on social relations or are in- ied process that interacts with the lifeworld of the consumer. In our tertwined with those relations. Due to the public and distributed na- findings, which are visually represented in Figure 1, we explore the ture of the story world, consumers regularly report interacting with experience of narrative transportation with geo-media, along with strangers also immersed in the Pokémon GO narrative. Yet, consum- the factors that support it. ers’ most consistent and profound social experiences are with those they already know: co-workers playing with each other outside of Dual-Move Narrative Transportation office hours; old friendships, forged over past Pokémon games, be- Geo-media enable a form of dual-move transportation: the nar- ing awakened through play after periods of dormancy; close friends rative moves into the physical lifeworld of the consumer, just as the altering the rhythm and routines of their relations with Pokémon- consumer moves into the story world. We find that consumers speak related trips (Quote 3.1). In this way, the story narrative structures of story world elements and physical world entities interchangeably social relations, providing people with a reason to be together as well as if they co-exist with each other in the same reality. They visit story as a social script to follow. world Pokéstops and Pokémon gyms, just as they would visit bus In turn, these social relations validate the locative content and stops or brick-and-mortar gyms (Quote 1.1). They imbue the Poké- make it seem more real and appealing. Acquaintances and strang- mon monsters with agentic qualities, describing how they live out ers, alike, talk about story world elements, visit Pokéstops and hunt in the world as independent entities that possess the will and ability Pokémon, and share pictures of their captures; they are creating to evade capture (Quote 1.2). Consumers are deeply immersed in shared knowledge, participating in rituals, and enacting norms and the story and internalize the role of Pokémon trainers as they go on values. In short, they are constructing a culture that recognizes the organized day long ‘hunts,’ as well as plan their schedule and travel existence of a hybridized narrative-physical reality. In this culture, routes around their participation in the narrative. some consumers even project social qualities on to the story world These attitudes and behaviors provide compelling evidence that characters themselves, imagining that Pokémon have social rela- consumers experience transportation into the Pokémon story world, tions with humans, pets, and other Pokémon (Quote 3.2). Taken as even as it is integrated in their own physical world. This dual move a whole, these findings suggest that geo-media storytelling can have stands in contrast to traditional conceptualizations of narrative trans- powerful impact on social relations and the experience of narrative portation in which the consumer, only, is cognitively transplanted transportation. into a separate, imaginary world. The question with transportation through geo-media becomes: how do stories come to be stabilized in Validated Imagery & Embodied Transportation Practices the physical world and accepted as a part of everyday life? Prior work on narrative transportation highlights the importance of story elements such as character and plot, and those cannot be un- Narrative Emplacement in the Physical World derestimated in the context of geo-media either. Yet, as previously Geo-media facilitated stories are spatialized. The characters are noted, the emplaced nature of geo-mediated stories in the real world in one’s home and the plot plays out around the consumer amidst also necessitates that these story elements be validated both spatially day-to-day activities. These are not the ethereal worlds of fiction- and socially through, for example, thoughtful integration into the al books and movies. We find that perceptions of, and movement surrounding context and social relations. Consumers, as story inter- through, space is shaped by this location-based story content (Quote preters and co-creators, play a critical role in this process, imagining 2.1). Consumers seek out and express excitement about areas that and discussing undeveloped aspects of the story (e.g. the existence of feature a high concentration of Pokéstops, or that are rumored to be Pokémon nests and relations between Pokémon monsters and pets). home to rare Pokémon. They also make pilgrimages to new places Another way that transportation differs with geo-media is that and move deliberately (e.g. zig-zagging, travelling in loops, etc.) it is an embodied practice: the story character’s body and the con- through space as a result of this narrative content. The impact of sumer’s body are one and the same, travelling around the hybridized locative content on space is represented by the dotted arrow pointing physical-story world. If a consumer wants to experience a particular towards ‘space’ in Figure 1. narrative thread, he or she needs to physically access it. And some Consumers also engage in efforts to cognitively validate the aspects of a story – those set in an ocean or in a foreign country – status of the emplaced narrative in the physical world (solid arrow can be physically, socially, monetarily hard to access (Quote 4.1). pointing away from ‘space’ in Figure 1). They identify pre-existing It is not as simple as turning the page or pressing play. Consumers Advances in Consumer Research (Volume 45) / 871 go on hunts that leave them hot and sweaty, they get lost, and they Graham, Mark, Matthew Zook, and Andrew Boulton (2012), avoid particular places because embodied narrative transportation is “Augmented Reality in Urban Places: Contested Content and accompanied by real and felt risks (Quote 4.2). the Duplicity of Code,” Transactions of the Institute of British Geographers, 38 (3), 464-479. Discussion Green, Melanie C., and Keenan M. Jenkins (2014), “Interactive In this research, we explored how to conceptualize narrative Narratives: Processes and Outcomes in User‐Directed Stories,” transportation when a story is told using geo-media, creating a mixed Journal of Communication, 64 (3), 479-500. reality through emplacing a story world within the story receiver’s Green, Melanie C., and Timothy C. Brock (2002), “In the Mind’s world of lived experience. We have found that narrative transporta- Eye: Transportation-Imagery Model of Narrative Persuasion,” tion, in this case, involves more than just creating mental imagery to in Narrative Impact: Social and Cognitive Foundations., ed. access a separate world. It involves embodied practices and mental Melanie C. Green, Jeffrey J. Strange, and Timothy C. Brock, acts of validation that create and stabilize a mixed reality in which Mahwah, NJ: Erlbaum, 315–41. the story world “feels real” as it is seamlessly emplaced in the physi- Green, Melanie C., Timothy C. Brock, and Geoff F. Kaufman cal world. Narrative transportation in mixed reality thus requires re- (2004), “Understanding Media Enjoyment: The Role of searchers and marketers to rethink the environment in which a story Transportation into Narrative Worlds,” Communication is consumed, as well as how the story is produced. Theory, 14 (4), 311–27. Narrative transportation research, due to its focus on laboratory Kuzmičová, Anežka (2015), “Does It Matter Where You Read? studies, typically downplays the role of the environment. If consid- Situating Narrative in Physical Environment,” Communication ered at all, the environment is conceptualized as a locus of distrac- Theory, 26 (3), 1-19. tion, for example when a crying baby in a movie theater attracts at- Lapenta, Francesco (2014), “Geomedia: On Location-based Media, tention to the real world, rather than the story world, and thus makes The Changing Status of Collective Image Production and the transportation difficult or impossible (Green et al. 2004). However, Emergence of Social Navigation Systems.” Visual Studies, 26 even in more traditional media, the environment can cross-fertilize (1), 14-24. mental imagery, for example when someone reads Heart of Dark- Leszczynski, Agnieszka (2015), “Spatial Media/tion,” Progress in ness, a story set on a boat traversing the Congo River, next to a bub- Human Geography, 39 (6), 729-751. bling stream. This type of environmental propping of mental imagery Liao, T., and L. Humphreys (2014), “Layar-ed Places: Using (Kuzmičová 2015), while typically ignored in narrative transporta- Mobile Augmented Reality to Tactically Reengage, tion research, becomes especially important for people and brands Reproduce, and Reappropriate Public Space,” New Media & who wish to use geo-media to tell their stories. Society, 1-18. Muniz, Albert M., and Thomas C. O’Guinn (2001), “Brand REFERENCES Community,” Journal of Consumer Research, 27 (4), 412–32. Ahn, Sun Joo, Amanda Minh Tran Le, and Jeremy Bailenson Phillips, Barbara J., and Edward F. McQuarrie (2010), “Narrative (2013), “The Effect of Embodied Experiences on Self-other and Persuasion in Fashion Advertising,” Journal of Consumer Merging, Attitude, and Helping behavior,” Media Psychology, Research, 37 (3), 368–92. 16(1), 7-38. Scholz, Joachim, and Andrew N. Smith (2016), “Augmented Biocca, Frank (2002), “The Evolution of Interactive Media,” Reality: Designing Immersive Experiences that Maximize Narrative Impact. Social and Cognitive Foundations, 97-130. Consumer Engagement,” Business Horizons, 59 (2), 149-161. Chess, Shira (2014), “Augmented Regionalism: Ingress Van Laer, Tom, and Ko de Ruyter (2010), “In Stories We Trust: As Geomediated Gaming Narrative,” Information, How Narrative Apologies Provide Cover for Competitive Communication & Society, 17 (9), 1105-1117. Vulnerability after Integrity-Violating Blog Posts,” Elson, Malte, Johannes Breuer, James D. Ivory, and Thorsten International Journal of Research in Marketing, 27 (2), Quandt (2014), “More Than Stories with Buttons: Narrative, 164–74. Mechanics, and Context as Determinants of Player Experience Van Laer, Tom, Ko De Ruyter, Luca M. Visconti, and Martin in Digital Games.” Journal of Communication, 64 (3), 521- Wetzels (2014), “The Extended Transportation-imagery 542. Model: A Meta-analysis of the Antecedents and Consequences Farman, J. (2012), Mobile Interface Theory, New York: Routledge. of Consumers’ Narrative Transportation,” Journal of Gerrig, Richard J. (1993), Experiencing Narrative Worlds: On the Consumer Research, 40 (5), 797-817. Psychological Activities of Reading, New Haven, CT: Yale University Press. Does Pulling Together Lead to Falling Apart? The Self-Regulatory Consequences of Cooperative Orientations for the Self-Reliant Ainslie Schultz, Providence College, USA Cait Lamberton, University of Pittsburgh, USA Jesper Nielsen, University of Arizona, USA

EXTENDED ABSTRACT were told the matrices contained two numbers summing to 10, and to mark “got it” when they solved the matrix (Mazar, Amir, and Ariely To be successful you have to be selfish, or else you never achieve. 2008). Since participants did not need to indicate which two numbers And once you get to your highest level, then you have to be summed to ten, participants could cheat and win; however, 10 of the unselfish. – Michael Jordan 20 matrices did not contain two numbers summing to 10. This allowed us to measure dishonest behavior by the number of unsolvable Many successful individuals often attribute their personal suc- matrices participants dishonestly checked as solved. cess to their own self-reliance and individual effort in training and Contrasts show that individuals primed with self-reliance and practice. As these individuals advance to managers who oversee oth- asked to cooperate showed more unethical behavior than those asked ers’ work, for example, or team captains, continued success often to compete individually (Mcooperation, self-reliance = 4.27; Mcompetition, self-reliance requires collaboration. Essentially, they must put their “selfish in- = 2.63; F(1, 168) = 3.87, p = .05), demonstrating that for individuals terests” aside and navigate the opinions and capabilities of others in primed with self-reliance, participating in a cooperative (vs. com- order achieve. In these situations, the question arises: what happens petitive individual) task led to more unethical behavior. when self-reliant individuals are placed in situations that depend on Experiment 2 uses a novel online shopping task to create a cooperation and others’ performance? “real” cooperative task. 174 undergraduates took part in a lab study. Past research shows cooperation can help individuals and or- Participants first completed the writing task from Experiment 1 to ganizations reach their goals (Johnson and Johnson 1986). Less prime self-reliance (vs. control). Next, participants were assigned a understood, though, are the consequences for those involved in co- partner or competitor and either cooperated with or competed against operation. Research has proposed that cooperation triggers personal their partner for five minutes to earn the highest score on the online costs for collaborators since cooperation requires restraining selfish search task. Participants received a shopping list of products and urges (Kocher et al. 2017), but support for this hypothesis is mixed were to find those items on the retailer’s website using only their with findings in support of this relationship (Kocher et al. 2017; mouse, no keyboard. Last, participants completed a series of four Verkoeijen and Bouwmeester 2014) and in conflict (e.g. Bear and anagrams—but unbeknownst to participants, the fourth anagram was Rand 2016; Lotito, Migheli, and Ortona 2013). One explanation for unsolvable (Baumeister et al. 1998). The time spent on the anagram this inconsistency is that not all individuals may experience coopera- served as our measure of restraint. tion’s depleting effects. Our research introduces an important moder- Individuals who cooperated and were primed with self-reliance ator, self-reliance, that we believe may help explain why cooperation persisted for less time than those who competed individually (M sometimes appears to require the expenditure of self-regulatory re- coop- eration, self-reliance = 101.54 seconds; Mcompetition, self-reliance = 143.20 seconds; sources, while at other times, self-regulatory resources remain intact. F(1, 170) = 4.23, p = .04), demonstrating that for individuals primed We employ ego depletion theory (Baumeister et al. 1998) and with self-reliance, participating in a cooperative (vs. competitive in- suggest that whether cooperation depletes an individual’s self-con- dividual) task led to less persistence on a subsequent task. trol depends on that individual’s level of self-reliance. Since the self- Experiment 3 tests self-regulation as the underlying mechanism reliant believe they can, and prefer to, succeed through their own and demonstrates that declines in self-regulation trigger these ef- efforts, ego depletion theory would suggest that for the self-reliant, fects. 172 participants were recruited on M-Turk. The experiment cooperating may be taxing. Cooperation requires relinquishing con- consisted of a between subjects design with one manipulated factor trol to others, switching mindsets, navigating other perspectives, and (cooperation or individual competition) and one measured variable acting against one’s preferences—behaviors that, for the self-reliant, (trait self-reliance). Participants first saw the same cooperative and likely draw on self-regulation and force counter-attitudinal behaviors competitive individual manipulations from Experiment 1 and after, (Baumeister et al. 1998). the number-summing task from Experiment 1, which served as our Experiment 1 predicts that individuals primed with self-reliance measure of restraint toward dishonest behavior. Last, participants (vs. no-prime control condition) show more unethical behavior, a completed a filler task, Triandis and Gelfand’s (1998) self-reliance measure of self-control, when cooperating versus competing indi- scale, and Tangney, Baumeister, and Boone’s (2004) self-control vidually. 172 undergraduates participated in an online study. Par- scale. ticipants first completed a writing task to prime self-reliance (vs. Analysis shows a significant two-way interaction of the cooper- control). In the self-reliance condition, participants wrote about an ative/individual competitive prime and self-reliance and a significant experience in their life when they felt self-reliant, while in the control three-way interaction of cooperative/individual competitive prime, condition, participants wrote about any experience in their life. self-reliance, and self-control. The same pattern of results from the Participants then saw a string of letters (“hdskltiwoertbca”) previous studies emerged among lower self-control individuals. and had three minutes to make as many words as possible out of Here, low self-control, high self-reliant individuals engaged in more the letters (e.g. ski). Participants believed that they were either (1) unethical behavior under cooperation as opposed to individual com- cooperating with others online (cooperation condition), or (2) competition (Effect = 1.92, t = 2.59, p = .01). The effect did not emerge peting against other individuals (individual competition condition). among those higher in self-control. Participants in both conditions were entered into a lottery based on Overall, our results show that whether cooperation requires their performance. For the dependent variable, participants received self-control depends on an individual’s level of self-reliance. We find 20 number matrices, containing 12, 3-digit numbers (e.g. 2.17) and that self-reliant individuals show diminished self-control when co-

Advances in Consumer Research 872 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 873 operating as opposed to competing individually. Our results delve Lotito, Gianna, Matteo Migheli, and Guido Ortona (2013), “Is deeper into why pulling together may lead to falling apart for self- cooperation instinctive? Evidence from the response times in a reliant individuals. public goods game,” Journal of Bioeconomics, 15(2), 123-33. Mazar, Nina, On Amir, and Dan Ariely (2008), “The dishonesty of REFERENCES honest people: A theory of self-concept maintenance,” Journal Baumeister, Roy F., Ellen Bratslavsky, Mark Muraven, and Dianne of Marketing Research, 45(6), 633-44. M. Tice (1998), “Ego depletion: Is the active self a limited Tangney, June P., Roy F. Baumeister, and Angie Luzio Boone resource?” Journal of Personality and Social Psychology, (2004), “High self‐control predicts good adjustment, less 74(5), 1252-65. pathology, better grades, and interpersonal success,” Journal Bear, Adam and David G. Rand (2016), “Intuition, deliberation, of Personality, 72(2), 271-324. and the evolution of cooperation,” Proceedings of the National Triandis, Harry C., and Michele J. Gelfand (1998), “Converging Academy of Sciences, 113(4), 936-41. measurement of horizontal and vertical individualism and Johnson, Roger T., and David W. Johnson (1986), “Cooperative collectivism,” Journal of Personality and Social Psychology, learning in the science classroom,” Science and Children, 74(1), 118-28. 24(2), 31-2. Verkoeijen, Peter P. J. L., & Bouwmeester, S. (2014), “Does Kocher, Martin G., Peter Martinsson, Kristian Ove R. Myrseth, and intuition cause cooperation?” PloS One, 9(5), e96654. Conny E. Wollbrant (2017), “Strong, bold, and kind: Self- control and cooperation in social dilemmas,” Experimental Economics, 20(1), 44-69. A Regulatory Focus Approach to Consumers’ Moral Decision Making Maria Schwabe, Friedrich-Schiller-University of Jena, Germany David B. Dose, Friedrich-Schiller-University of Jena, Germany Gianfranco Walsh, Friedrich-Schiller-University of Jena, Germany

EXTENDED ABSTRACT nipulating subsequent decisions of the same domain we first let par- This research demonstrates that consumers’ regulatory focus in- ticipants choose from nine convenience goods (e.g., coffee, potato fluences moral self-regulating consumption. Promotion-focused con- chips) that they would like to buy (Mazar and Zhong 2010). In the sumers’ moral deeds provide them with a license for subsequent con- more moral condition, participants were shown seven green and two sumption behavior that is more immoral, and more immoral deeds conventional convenience goods; in the more immoral condition, can be cleansed by successively more moral consumption decisions. seven conventional and two green convenience goods appeared. Af- Prevention-focused consumers repeat their past moral consumption terwards, they indicated their relative preference for organic versus decisions, irrespective of the moral valence of previous decisions conventional coffee as dependent measure. In the different domain though. We further show that these effects are contingent on the do- condition, participants chose from a set of nine (moral: 7 green, 2 main congruency of the decisions and the efficiency of a more moral conventional; immoral: 7 conventional, 2 green) vouchers redeem- option. able for green versus conventional clothes (e.g., T-shirts, jeans). The Recent literature has shown that consumers’ moral decisions results indicate that promotion-focused participants show balancing largely depend on their behavioral history, though research into patterns only for subsequent consumption situations in different be- how previous decisions influence consumers’ subsequent behavior havioral domains. For subsequent decisions in the same behavioral in moral trade-off situations offers contradictory results: On the one domain they mimic their past decisions irrespective of the behavior’s hand, consumers might persist in their moral history, leading to con- moral stance. Prevention-focused consumers, however, show consis- sistent behavior (e.g., Gino, Norton, and Ariely 2010; Karmarkar and tency in their consumption behavior in the same as well as in differ- Bollinger 2015; Zhang et al. 2014). On the other hand, consumers ent consumption domains. may deviate from the moral stance reflected in their past decisions, Study 3 manipulated the morality of the first decision, con- such that they balance their moral and immoral deeds (Gneezy et sumers’ regulatory focus, and the charity’s efficiency, using a 3 al. 2012; Krishna 2011). Presented with these opposing theoretical (first decision: moral vs. immoral vs. control) ´ 2 (regulatory focus: viewpoints, more research is needed to determine when consumers prevention vs. promotion) ´ 2 (charity efficiency: high vs. low) be- are motivated to balance their consumption decisions and when they tween-subjects design with 372 students and university employees tend to be consistent (Mullen and Monin 2016). Drawing on regula- as participants. The dependent variable was the choice in a moral tory focus theory, we demonstrate that consumers’ regulatory focus trade-off, that is, the choice between keeping the reward in a cus- influences moral self-regulating consumption. tomer referral program for themselves and donating the money to Study 1 was a 3 (first decision: moral vs. immoral vs. control) charity. Regulatory focus was manipulated as in study 1, while we ´ 2 (regulatory focus: prevention vs. promotion) between-subjects used the scenario with nine different convenience goods from study 2 experiment with a sample of 188 students and university staff. Regu- for manipulating the morality of the first decision. Charity efficiency latory focus was manipulated by asking participants to write about was manipulated by indicating that the charity received a two-star their hopes/aspirations (promotion focus) or obligations/duties (pre- (low efficiency) or four-star (high efficiency) rating by an indepen- vention focus) (Freitas and Higgins 2002). To manipulate the moral- dent institution (Winterich and Barone 2011). The results show that ity of the first decision participants wrote about a situation where charity efficiency has no influence on the moral consumption behav- they helped someone unrewarded (moral) or were mean without any ior of promotion-focused participants. The authors presume that this particular reason (immoral) (Conway and Peetz 2012). In all studies is because promotion-focused consumers concentrate on the conse- the control condition consisted of an unrelated scrambled sentence quences of their behavior for their own moral self-perception, where- task. After the writing tasks, participants indicated their relative pref- as the degree to which their moral decision benefits others may have erence for organic versus conventional coffee which was the depen- relatively less importance. However, prevention-focused consumers dent measure. Interestingly, the results reveal that promotion-focused refrain from mimicking the moral stance of their past behavior if the participants who had previously recalled moral decisions exhibited moral alternative is inefficient. Instead, they likely choose the more licensing behavioral patterns, indicating a lower subsequent prefer- immoral option, irrespective of their behavior in subsequent situa- ence for the organic coffee. Participants in the promotion focus conditions. An inefficient moral choice option creates a situation in which tion who remembered past immoral decisions instead tried to “clean the more moral alternative also is more risky, in that the donated up” their behavior by preferring organic over conventional coffee. money may not reach the intended recipient and thus would be lost. An induced prevention focus, however, causes consumers to repeat Hence, prevention-focused consumers may opt for the more immoral their past behavior which is consistent with past research (Zhang et alternative, irrespective of their preceding decisions. al. 2014). Hence, participants with a prevention focus who recalled In sum, the findings show that a consumer’s inherent or situ- moral decisions preferred the organic coffee over the conventional ationally induced regulatory focus influences whether he or she alternative. If they confronted a prior immoral decision, prevention- regulates or repeats moral consumption decisions. These effects, focused participants indicated less preference for the organic coffee. however, are contingent on boundary conditions such as the domain In study 2, we measured consumers’ chronic regulatory fo- congruency of the decisions studied and the efficiency of the more cus and manipulated the morality of the first decision, as well as moral option. Thus, our research improves understanding regarding the domain congruence of the first and second decisions, with a 2 the dynamics of consumers’ moral self-regulation. (first decision: moral vs. immoral) ´ 2 (domain: same vs. different) between-subjects design and a morally neutral control group. The sample consisted of 291 students and university employees. For ma-

Advances in Consumer Research 874 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 875 REFERENCES Krishna, Aradhna (2011), “Can supporting a cause decrease Conway, Paul and Johanna Peetz (2012), “When does feeling moral donations and happiness? The cause marketing paradox,” actually make you a better person? Conceptual abstraction Journal of Consumer Psychology, 21(3), 338-45. moderates whether past moral deeds motivate consistency or Mazar, Nina and Chen-Bo Zhong (2010), “Do green products make compensatory behavior,” Personality and Social Psychology us better people?” Psychological Science, 21(4), 494-98. Bulletin, 38(7), 907-19. Mullen, Elizabeth and Benoît Monin (2016), “Consistency versus Freitas, Antonio L. and E. Tory Higgins (2002), “Enjoying goal- licensing effects of past moral behavior,” Annual Review of directed action: The role of regulatory fit,”Psychological Psychology, 67(1), 363-85. Science, 13(1), 1-6. Winterich, Karen Page and Michael J. Barone (2011), “Warm Gino, Francesca, Michael I. Norton, and Dan Ariely (2010), “The glow or cold, hard cash? Social identity effects on consumer counterfeit self the deceptive costs of faking it,” Psychological choice for donation versus discount promotions,” Journal of Science, 21(5), 712-20. Marketing Research, 48(5), 855-68. Gneezy, Ayelet, Alex Imas, Amber Brown, Leif D. Nelson, and Zhang, Shu, James FM Cornwell, and E. Tory Higgins (2014), Michael I. Norton (2012), “Paying to be nice: Consistency “Repeating the past: Prevention focus motivates repetition, and costly prosocial behavior,” Management Science, 58(1), even for unethical decisions,” Psychological Science, 25(1), 179-87. 179-87. Karmarkar, Uma R. and Bryan Bollinger (2015), “BYOB: How bringing your own shopping bags leads to treating yourself and the environment,” Journal of Marketing, 79(4), 1-15. The Chill of the Moment: Emotions and Pro-Environmental Behavior Daniel Schwartz, University of Chile, Chile George Loewenstein, Carnegie Mellon University, USA

EXTENDED ABSTRACT are more affected by sadness-evoking ads and, therefore, by the effect In the internet age, visual emotional appeals have gained a of a time delay. This result was replicated in the following studies. broader audience, and hence attracted great attention from marketing Study 2 examines an informative intervention in which partici- researchers, as many short videos or ads are shared online, gener- pants (N = 520) are warned of the effect of emotions on behavior ally watched only once. These videos, which often request immedi- (“When time has passed since people have watched the video, and ate action (e.g., a video about world hunger followed by a donation the emotions that the video evokes have cooled down, people are request), have become an important tool to promote action on public less likely to donate, and when they do donate, they donate smaller policy issues. One important question is whether these social-mar- amounts”). We also examined whether an active choice may increase keting emotional materials will motivate the same action after a time donations (alone or with the warning) by asking participants whether delay once emotions have cooled off. This is especially important they would like to postpone their donation decision. This study used for environmental policies as many actions (e.g., purchasing a green the same sadness-evoking video from study 1, with a one-day time product) are likely to be taken only after a time delay. window between parts. We found that participants donated 45% Climate change may be an appropriate target for interventions more right after watching the emotion-evoking video than when using emotional ads because it has almost all the hallmarks of an asked to make a donation the day after (p < 0.01). When presented issue about which people are likely to experience scarce moments with the possibility of postponement, participants offered a donation of emotional arousal to motivate pro-environmental action (Loew- 34% higher than those who could only donate the next day, p = 0.03, enstein and Schwartz 2010). Prior research has addressed the effect but very similar to participants who could only donate right after of emotion-evoking stimuli on environmental issues, focusing on in- watching the video, p = 0.56. This suggests that choosing may be a tentions or attitudinal measures (Hartmann et al. 2016; Leiserowitz useful strategy to avoid the diminution of donations following a de- 2004; Lowe et al. 2006; Matthes, Wonneberger, and Schmuck 2014). lay. We found that explicitly warning participants that emotions will However, studies that only provide hypothetical opportunities for diminish and affect their donations has no effectp ( = 0.63). taking action may actually understate the effect of emotion-evoking Study 3 adds a non-binding commitment just after participants stimuli by not providing real opportunities to dispel negative emo- (N = 476) watched the sadness-evoking video (“We would like to tions (Witte and Allen 2000). Furthermore, in most prior research, ask you how much you would like to donate in the second part of attitudes and intentions have been measured immediately follow- the study”). We found that adding a non-binding donation caused ing the emotion-evoking stimulus, raising the question of whether participants to donate, one day later, an amount very similar to what changes in attitudes or behavioral intentions would continue over they proposed the previous day (r = 0.96). This high correlation re- time. We tackle these issues by measuring actual behavior immedi- sulted in higher donations on the second day of the study compared ately following, or with a delay after, an emotion-inducing stimulus. to participants who watched the same video but did not indicate any We introduce two interventions to offset the effect of the time intention the previous day (30.8% difference; p < 0.01). delay using tools that expand the literature of social marketing. First, In summary, these studies reveal that sadness-evoking videos as practitioners may assume that warning people about the conse- induce greater donations to an environmental organization than quences of time delay is an effective tool, we examine the effect of non-affective videos, but this effect disappears once emotions have warning that emotions affect behavior, but this effect diminishes over cooled off after a delay. In addition, cautioning people that emotions, time. Second, we examine whether intentions, through non-binding and their effects on behavior, cool off is not effective unless people commitments adopted at a moment of emotional arousal, would provide a non-binding commitment just after watching the affective translate into actions after emotions have cooled off. ad so that they can translate their initial intention into action after Study 1 examines the differential impact of sadness-evoking and the time delay. Our results help explain why emotion-evoking ads non-emotional videos, adding a time delay between videos and par- that seek to promote pro-environmental behavior, such as reducing ticipants’ decisions. Participants (N = 687) were randomly assigned energy use, often fail to produce sustained behavioral change, and to watch either a sadness-evoking video or a non-emotional video, they suggest that if a sustained response is sought, behavioral com- both previously pretested. Participants were randomly assigned to mitments may need to be elicited in moments of high emotion. donate all or part of an announced $30 prize to the World Wildlife Fund (WWF), either immediately or one hour after they watched REFERENCES the video. We found that those who watched the sadness-evoking Hartmann, Patrick, Vanessa Apaolaza, Clare D’Souza, Jose video offered to donate 25% more money right after they watched M. Barrutia, and Carmen Echebarria (2016), “Promoting the video than one hour later, p = 0.03. Time had no impact on those Renewable Energy Adoption: Environmental Knowledge vs. who watched the non-emotional video, p = 0.57. Participants who Fear Appeals,” Springer, Cham, 359–67. watched the emotion-evoking video offered to donate more money Leiserowitz, AA (2004), “Day After Tomorrow: Study of Climate (30%) than participants who watched the non-emotional video when Change Risk Perception,” Environment: Science and Policy for donations were made just after watching the videos, p < 0.01. How- Sustainable Development, 46(9), 22–39. ever, one hour after watching the videos, there was no difference in Loewenstein, G. and Daniel Schwartz (2010), “Nothing to Fear donations, p = 0.89. Reported sadness mediated the decrease in do- but a Lack of Fear: Climate Change and the Fear Deficit,”G8 nations, after the delay, by participants who watched the emotion- Magazine, 60–62. evoking video. Based on ELM theory (Petty and Cacioppo 1990), we found that people with lower (vs. higher) concern for global warming

Lowe, T., K. Brown, S. Dessai, M. F. Doria, K. Haynes, and K. Petty, Richard E. and John T. Cacioppo (1990), “Involvement Vincent (2006), “Does Tomorrow Ever Come? Disaster and Persuasion : Tradition Versus Integration,” American Narrative and Public Perceptions of Climate Change,” Public Pschyological Association, 107(3), 367–74. Understanding of Science, 15(4), 435–57. Witte, Kim and M Allen (2000), “A Meta-Analysis of Fear Appeals: Matthes, Jörg, Anke Wonneberger, and Desirée Schmuck (2014), Implications for Effective Public Health Campaigns.,”Health “Consumers’ Green Involvement and the Persuasive Effects education & behavior : the official publication of the Society of Emotional versus Functional Ads,” Journal of Business for Public Health Education, 27(5), 591–615. Research, 67(9), 1885–93. Will It Taste Better If You Think about What You Are Eating? Cultural Differences in Food-Ingredient Information Seeking Hao Shen, Chinese University of Hong Kong, China Jun Pang, Renmin University of China, China

EXTENDED ABSTRACT = 18.30, p < .001). However, this difference was mitigated for Chinese When consumers choose foods, some may prefer items for which participants (M = 5.27 vs. 4.95, F (1, 193) = 1.31, p > .25). The two- they are offered specific information about the ingredients constituting way interaction between culture and ingredient information was also the items they will be eating, whereas others may not form preferences significant (F (1, 193) = 4.81, p = .03), which was independent of food based on such information. Thus, an important question regarding food type (F < 1.0). marketing arises: Should food be marketed by providing specific infor- Study 3 used a 2 (culture: Chinese vs. American) × 2 (processing mation about ingredients to consumers? style: overall impression vs. ingredients focus) between-subjects de- The answer to this question may depend on the consumer’s cultur- sign. Participants saw a picture of seafood pasta and were either asked al background. In this research, we propose that Chinese versus Ameri- to form an overall impression of food or write down each ingredient can consumers may have different lay beliefs about how different foods in this dish. We then asked them to indicate the tastiness of the food. come together to form a meal, affecting norms for preparing and serving The two-way interaction between culture and processing style was sig- foods. Due to Confucian teaching and traditions, Chinese people tend nificant (F(1, 140) = 6.10, p = .015). In particular, Americans evaluated to believe in the philosophy of yin and yang, such that apparently op- pasta more favorably if they wrote down each ingredient than if they posite forces are complementary and need to be balanced. This belief gave an overall impression (M = .65 vs. -.04, F(1, 140) = 3.40, p = .07). permeates the norms and values associated with foods and eating. When However, the reverse was true for Chinese people, though the difference preparing foods, Chinese people want to balance yin foods (e.g., fruits was marginally significant (M = .87 vs. 1.45, F(1, 140) = 2.71, p = .10). and vegetables) and yang foods (e.g., beef and pork). Chinese food is an Study 4 used a 2 (culture: Chinese vs. American) × 2 (food pre- art of mixture that balances color, flavor, and texture. Chinese cooking sentation: separate vs. mixed) between-subjects design with the type involves looking at the combination of the ingredients as well as pay- of food as a within-subject factor. Participants saw a picture of a beef ing particular attention to the complex process and equipment involved. noodle dish and a beef burger, and the order of these pictures was coun- China has hundreds of cooking methods. Good cooking depends on the terbalanced. In the separate condition, each ingredient was displayed blending of various ingredients and condiments rather than the taste of separately in both pictures. In the mixed condition, ingredients were the individual elements. Chinese people believe that various ingredi- presented mixed together in both pictures. A 2 × 2 ANOVA of culture ents result in the overall harmony of flavors (Sundararajan 2015). For and display on food evaluation revealed a significant interaction (F example, when preparing a dish of “Beef with mushroom and bamboo (1,151) = 9.63, p < .01), which was independent of food type. More shoots,” Chinese people put all ingredients into a pan and deep-fry them specifically, separated (vs. mixed) presentation increased food evalua- repeatedly, to mix the flavors of all the ingredients. In the pursuit of bal- tions for Americans (Mseparate = 1.09, Mmixed = .49, F (1,151) = 4.90, p = ance and moderation, Chinese people do not perceive each ingredient to .03) but decreased food evaluations for Chinese participants (Mseparate = be a stand-alone, independent element of the meal, but rather consider .14, Mmixed = .70, F (1,151) = 4.73, p = .03). the dish that has been formed to be a new, different entity—separate In sum, our research contributes to literature examining consum- from its component parts. The Judeo-Christian traditions of Americans, ers’ food perceptions, an area of increasing importance in consumer on the other hand, eschew balance in favor of extreme stances (Briley, behavior research (Bagchi and Block, 2011; Bublitz, Peracchio, and Morris, and Simonson, 2000; Spencer-Rodgers et al., 2009). Consis- Block, 2010; Wansink, 2015). Further, we address the influence of tent with this belief, Americans tend to regard each ingredient as inde- culture, showing that Chinese consumers may be less interested than pendent, and focus on how each element can contribute (Sundararajan, American consumers in information about individual food ingredients. 2015). Based on these considerations, we propose that Americans empha- REFERENCE size individual ingredients more than the Chinese do and therefore tend Bagchi, Rajesh, and Lauren G. Block (2011), “Chocolate Cake to be more interested in detailed descriptions of each ingredient when Please! Why Do Consumers Indulge More When It Feels More they consider foods. Study 1 provided initial evidence for our predic- Expensive?” Journal of Public Policy & Marketing, 30(2), tion. We asked participants to imagine going to a restaurant for a dinner 294-306. and finding the restaurant has two versions of the menu available: one Briley, Donnel A, Michael W Morris, and Itamar Simonson (2000), has the names of dishes with pictures, whereas the other has the names “Reasons as Carriers of Culture: Dynamic vs. Dispositional of dishes with detailed information about ingredients. American partici- Models of Cultural Influence on Decision Making,”Journal of pants (60.7%) were more likely than Chinese participants (31.7%) to Consumer Research, 27 (2), 157-78. choose the menu with ingredient information, (χ 2 (1) = 17.28, p <.001). Bublitz, Melissa G., Laura A. Peracchio, and Lauren G. Block Study 2 used a 2 (culture: American vs. Chinese) × 2 (food type: (2010), “Why Did I Eat That? Perspectives on Food Decision soup vs. pizza) × 2 (ingredient information: present vs. absent) between- Making and Dietary Restraint,” Journal of Consumer subjects design. We told participants we were interested in their choice Psychology, 20(3), 239-258. of foods. Then they were either exposed to a pair of soup dishes or a Spencer-Rodgers, J., Boucher, H. C., Mori, S. C., Wang, L., and pair of pizzas, depending on the condition. In addition, we manipulated Peng, K. (2009). The Dialectical Self-Concept: Contradiction, the availability of ingredient information. In one condition, both food Change, and Holism in East Asian Cultures,” Personality and options had no ingredient information. In the other condition, however, Social Psychology Bulletin, 35(1), 29-44. both options provided ingredient information. American participants re- Sundararajan, Louise (2015), Understanding Emotion in Chinese ported greater choice certainty when ingredient information was present Culture. Switzerland: Springer International Publishing. for food options than when it was absent (M = 5.57 vs. 4.37, F (1, 193)

EXTENDED ABSTRACT are incentive-comparable. In a typical experiment, participants were In the last decade, two lines of research on uncertainty have assigned to one of six conditions, which constitute a 3 (outcome: found their places in the literature: one by Gneezy, List, and Wu certain high (HI) vs. certain low (LO) vs. uncertain (UN)) x 2 (deci- (2006) documenting a strong uncertainty-aversion effect, and one by sion mode: in-prospect vs. in-process) between-participants design. Shen, Fishbach, and Hsee (2015) documenting a strong uncertainty- In one study, participants were given the opportunity to repeatedly loving effect. Each line is surprising by itself, and more surprising is guess the meaning of a foreign word, and were promised to earn the contrast between the two. This research aims to reconcile these $0.25 (LO), $0.50 (HI), or either $0.25 or $0.50 (UN) if they made two striking (and strikingly opposite) effects and build up a theoreti- the correct guess. The word was designed to be so difficult that no cal framework to understand responses to uncertainty in general. one could guess it correctly, so the DV was persistence. As for an in- Consumers have different responses to exogenous uncertainty, prospect decision, the participants, before starting the task, indicated that is, the out-of-personal-control kind of uncertainty. It appears how long they would persist. As for an in-process decision, the par- that most of the times, they are uncertainty averse (e.g., Arrow 1965; ticipants worked on the task and we observed how long they actually Holt and Laury 2002; Kahneman and Tversky 1979) and sometimes persisted. We found that for in-prospect decisions, the persistency uncertainty loving (e.g., Dhar, Gonzalez-Vallejo, and Soman 1995; pattern followed: HI > UN = LO, indicating not so strong uncertainty Goldsmith and Amir 2010; Mazar, Shampanier, and Ariely 2016). aversion and failing to replicate Gneezy et al. (2006), whereas for It is important to note that as long as the response to the uncertain in-process decisions the persistency pattern followed: UN > HI> LO, outcome is still bounded by the responses to its worst and best out- indicating strong uncertainty loving and succeeding in replicating comes for certain, both uncertainty aversion and loving here are still Shen et al. (2015). In another study, we found the strong uncertainty congruent to the dominance principle, that is, the certain incentive of aversion did occur in in prospect, but as documented by Yang, Vosge- a high value should dominate the uncertain incentive of the expected rau, and Lowenstein (2013), the uncertain outcome had to be framed value, and in turn, the uncertain discount should dominate the certain as a lottery. incentive of a low value. In other studies, we further replicated this in-prospect/ in- Anomalies do exist. As a violation of dominance against the process discrepancy on (a) ambiguity (when UN does not specify lower bound, Gneezy et al. (2006) demonstrated a strong uncertain- the probability of each outcome), (b) choices (when the behavioral ty-aversion effect. Research participants were willing to work even measurement was a discrete variable), and (c) pricing measurement less to earn a lottery ticket of an uncertain payment than its worst (when the behavioral measurement has sensible financial conse- possible payment. As a violation of dominance against the upper quences). In the last study, we directly tested the continuity hypoth- bound, Shen et al. (2015) demonstrated a strong uncertainty-loving esis by breaking an in-process decision into several in-prospect deci- effect. Research participants worked even harder to earn a bonus of sions and gathered supportive evidence. an uncertain magnitude than a bonus of a certain, larger magnitude. To the best of our knowledge, we are the first to reconcile the We describe both above effects as “strong,” because each is a viola- Gneezy effect and the Shen effect, and we are also the first to demon- tion of the dominance principle. strate both effects by manipulating a single factor in a single experi- In this research, we suggest that they actually do not happen at ment with the same decision context and on the same magnitude lev- the same time. Strong uncertainty-aversion occurs before one engag- el. This research does not only resolve a “conflict” in the literature es in the activity, or “in-prospect,” while strong uncertainty-loving but also greatly advances our understanding of how people behave occurs when one engages in the activity, or “in-process.” Thus, the under uncertainty in prospect and in process. mode in which decisions are made, in-prospect versus in-process, moderates the effect of this incentive uncertainty. Specifically, we REFERENCES hypothesize that strong uncertainty aversion occurs in prospect, Arrow, Kenneth J. (1965), “Aspects of the Theory of Risk Bearing,” while strong uncertainty loving occurs in process. Helsinki, Sweden: Yrjö Jahnssonin Sätiö. The in-prospect and in-process decisions vary on many psycho- Dhar, Sanjay K., Claudia Gonzalez-Vallejo, and Dilip Soman logical aspects (see Figure 1 for a summary of the literature review). (1995), “Brand Promotions as a Lottery,” Marketing Letters, However, all psychological differences have their root in one physi- 6(3), 221-33. cal feature: continuity. The in-prospect decision is often made on a Gneezy, Uri, John A. List, and George Wu (2006), “The Uncertainty single time point and hence can be seen as a static decision in a tem- Effect: When A Risky Prospect Is Valued Less Than Its Worst poral isolation. By contrast, the in-process decision is often made Possible Outcome,” Quarterly Journal of Economics, 121(4), over time and hence can be seen as a dynamic decision in a temporal 1283-1309. continuity. Therefore, we propose a “continuity hypothesis”: that the Goldsmith, Kelly and On Amir (2010), “Can Uncertainty Improve effect from a single decision with continuity (in-process decisions) Promotions?” Journal of Marketing Research, 47(6), 1070-77. differs from the combined effects from multiple decisions in isolation Holt, Charles A., and Susan K. Laury (2002), “Risk Aversion and (in-prospect decisions). Continuity adds extra utility into the process, Incentive Effects,”American Economic Review, 92(5), 1644- and this process utility is usually in the form of high arousal and 55. emotion. In the case of exogenous uncertainty, the extra utility can Kahneman, Daniel, and Amos Tversky (1979), “Prospect Theory: be in the form of excitement and engagement (Goldsmith and Amir An Analysis of Decision under Risk,” Econometrica, 47(2), 2010; Lee and Qiu 2009; Shen et al. 2015; Vosgerau, Wertenbroch, 263-92. and Carmon 2006; Wilson et al. 2005). Lee, Yih Hwai and Cheng Qiu (2009), “When Uncertainty Brings We test this framework in a series of four experiments. All ex- Pleasure: The Role of Prospect Imageability and Mental periments entail real consequences for the participants and hence Imagery,” Journal of Consumer Research, 36(4), 624-33.

Mažar, Nina, Kristina Shampanier, and Dan Ariely (2016), “When Wilson, Timothy D., David B. Centerbar, Deborah A. Kermer, and Retailing and Las Vegas Meet: Probabilistic Free Price Daniel T. Gilbert (2005), “The Pleasures of Uncertainty: Pro- Promotions,” Management Science. longing Positive Moods in Ways People Do Not Anticipate,” Shen, Luxi, Ayelet Fishbach, and Christopher K. Hsee (2015), Journal of Personality and Social Psychology, 88 (1), 5-21. “The Motivating- Uncertainty Effect: Uncertainty Increases Yang, Yang, Joachim Vosgerau, and George Loewenstein (2013), Resource Investment in the Process of Reward Pursuit,” “Framing Influences Willingness to Pay but Not Willingness to Journal of Consumer Research, 41, 1301-15. Accept,” Journal of Marketing Research, 50(6), 725-38. Vosgerau, Joachim, Klaus Wertenbroch, and Ziv Carmon (2006), “Indeterminacy and Live Television,” Journal of Consumer Research, 32 (4), 487-95. Embracing Experiential over Material Consumption: Thinking about Death Increases Consumer Preferences for Experiences Yoonji Shim, University of British Columbia, Canada Katherine White, University of British Columbia, Canada

EXTENDED ABSTRACT riential versus material options in response to mortality salience is the pursuit of meaningfulness. Specifically, we predict that people And a moment came that stopped me on a dime, will perceive experiences (vs. material goods) as more meaningful I went sky diving, following mortality salience threats and will demonstrate a prefer- I went rocky mountain climbing, ence for experiential options because of a desire to pursue meaning. I went 2.7 seconds on a bull name Fumanchu. In five studies, we investigate whether mortality salience leads And I loved deeper, people to prefer experiential over material consumption, using mul- And I spoke sweeter, tiple real-world field (Studies 1a and 1b) and experimental settings And I gave forgiveness I’ve been denying, (Study 2). We further examine whether the observed tendency is And he said someday I hope you get the chance, mediated by differences in perceived meaningfulness of experiential To live like you were dying. and material purchases (Study 3). Finally, we test the moderating role —”Live Like You Were Dying”, Tim McGraw (2004) of meaning-affirmation on the observed effect (Study 4). In this song, which was the number one country song on Bill- board in 2004, a man who is going to die soon, makes the most of Study 1a his last days by engaging in various experiential pursuits such as sky In Study 1a, we sought to provide a preliminary test of our pre- diving and mountain climbing. He further advises that people take dictions in a field study. Borrowing a method from a global art proj- the opportunity to live like they were dying because, in the face of ect, Before I Die (www.beforeidie.cc), we examined whether people his own impending death, he was finally able to live life to the fullest. express greater desires for engaging in experiences than acquiring Similar insights were portrayed by the film, The Bucket List (2007), material goods when thinking about their own death. We created a which depicts a story about two men who ended up sharing a hospital chalkboard wall, stenciled with a grid of the sentence “Before I die, room at a crossroads in their lives and set out on the adventure of a I want to ______.” on campus, and then asked people walking by lifetime to spend the time they have left doing everything they ever to fill in the blank (n = 227) (see Figure 1). Once the wall was full, wanted to experience before they die. we photographed wall and transcribed the responses for coding pur- Although these anecdotal examples of embracing experiences poses. Then, the responses were erased so that a new set of sentences in response to reminders of death may seem intuitive at first blush, could be written on the wall. All of the responses were classified into past research on mortality salience finds that people in fact tend to one of the two categories (experiential vs. material) by two trained be more materialistic in the face of thinking about their own death coders. We calculated the proportion of responses that reflected expe- (Arndt et al. 2004; Kasser and Sheldon 2000; Mandel and Heine riences and the proportion of responses that reflected material goods. 1999). In contrast to the research suggesting that people will con- As predicted, people expressed a greater desire for engaging in sume material goods under conditions of mortality salience, we ar- experiences (e.g., go skydiving, ski in the Alps) than for acquiring gue that, when people are given the choice between material and material goods (e.g., get a Ferrari, buy all Apple devices) in the face of their own death (Mexp = 97.4%, SD = .01; Mmat = 2.6%, SD = .37; experiential options, thinking about one’s own death will make 2 people prefer experiential over material consumption. We propose χ (1) = 203.63, p < .001). Using a real-world setting, Study 1a pro- that the mechanism underlying this effect is the tendency for experi- vided initial correlational evidence for the notion that mortality sa- ential rather than material options to fulfill one’s desire for meaning. lience is related to preferences for acquiring experiences rather than We argue that reminders of one’s own mortality will lead people to preferences for acquiring material goods. imbue their lives with a sense of meaning and this activated desire Study 1b for meaning will make people prefer more experiential rather than In Study 1b, we adopted a different real-life approach and iden- material consumption e. To our knowledge, no work has looked at tified consumer preferences for experiences over material goods the comparative preferences for experiential versus material options in response to thinking about one’s own death using a natural data in response to mortality salience. source: Pinterest (www.pinterest.com). Pinterest is an online content Our conceptual framework extends previous work on TMT sharing platform that allows members to pin images, videos, and (Greenberg et al. 1986), which demonstrates that reminders of one’s other objects to their pin boards for others to view. To test our pre- own death creates existential anxiety, leading to responses that allow diction in a real-world online setting, we explored what people have the individual to bolster their own cultural worldviews in ways that pinned on Pinterest by using a keyword phrase: “Before I die.”. A re- increase perceptions of a life with meaning (Rosenblatt et al. 1989). search assistant used Pinterest Quick Search by typing in a keyword Building upon this, we make a novel prediction that thinking about phrase “before I die” in the search box and randomly selected 200 one’s own death leads individuals to prefer experiential purchases pin boards, which displayed a total of 2964 pins. The pins were clas- (i.e., those made with the primary intention of acquiring a life experi- sified into one of the two categories (experiential vs. material) by two ence) over material purchases (i.e., those made with the primary in- trained coders. We calculated the number of experiences or material tention of acquiring a tangible good or material possession; e.g., Van goods out of the total number of pins. Boven and Gilovich 2003). We make this key prediction because ex- As anticipated, people posting a pin listed under “before I die” periential purchases are particularly associated with a person’s true expressed a greater desire for engaging in experiences (e.g., go zip self (Carter and Gilovich 2012; Schlegel et al. 2009), which serves lining, stay in an ice hotel) than acquiring material goods (e.g., own as an important source of meaning in life (King et al. 2006). Thus, a pearl necklace, have a luxurious bathroom) (M = 95.7%, SD = we suggest that the mechanism underlying the preference for expe- exp Advances in Consumer Research 881 Volume 45, ©2017 882 / Embracing Experiential over Material Consumption: Thinking about Death Increases Consumer Preferences for Experiences

2 .03; Mmat = 4.3%, SD = .12; χ (1) = 2481.43, p < .001). The results Results and Discussion of Study 1b provided supporting evidence for the impact of mortality A one-way ANOVA on the index of WTP revealed a signifi- salience on preferences for experiences over material goods. cant difference between the conditions (F(1,156) = 4.85, p < .05). As predicted, participants who focused on thoughts about their own Study 2 death indicated higher WTP for experiential than material purchases In Study 2, we sought more direct evidence regarding the (Mmortality = $348.52, SD = 766.72), compared to those who thought causal effect of mortality salience on preferences for experiential about grocery shopping (Mcontrol = $16.79, SD = 1101.02). over material consumption in a more controlled setting. Given that To test our proposed underlying mechanism, we first calculated thoughts about death produce a unique type of anxiety (Greenberg an index of perceived meaningfulness by subtracting the average et al. 1990), we predicted that participants who focus on thoughts perceived meaningfulness of the three material purchases from the about death only indicate higher WTP for experiential than material average perceived meaningfulness of the three experiential purchas- purchases, but this effect would not emerge for those who focus on es. We then conducted a mediation analysis (model 4; Hayes 2013) thoughts about social rejection, physical pain, or a neutral topic. with mortality as the independent variable, the index of WTP for ex- Method and Procedure periential and material purchases as the dependent variable, and the We randomly assigned participants (n = 392) to one of four index of perceived meaningfulness as the mediating variable. First, experimental conditions: 1) the mortality threat, 2) social rejection we found a marginally significant main effect of mortality salience threat, 3) physical threat, and 4) grocery shopping (control) condi- (vs. control) on the index of WTP (b = 278.61, SE = 150.97, t = tion, and manipulated the conditions using a writing task (Arndt et 1.85, p = .067), indicating that participants who thought about their al. 2002). After completing this task, participants were asked to list own death (vs. control) indicated higher WTP for experiential than three experiential and material purchases and indicate their willing- material purchases. Furthermore, there was a significant main effect ness to pay (WTP) for each of the purchases. of mortality salience (vs. control) on the index of perceived meaningfulness (b = .46, SE = .22, t = 2.08, p < .05), indicating that par- Results and Discussion ticipants in the mortality salience condition perceived experiences First, we calculated participants’ WTP for the experiential and (vs. material goods) as more meaningful (vs. control). Finally, the material purchases by subtracting the average WTP of the three procedures generated a 95% confidence interval around the indirect material purchases from the average WTP of the three experiential effect with zero falling outside the confidence interval (b = 53.10, purchases. A one-way ANOVA on the index of WTP revealed a sig- SE = 38.51, 95% CI = [7.9077 to 165.5168]), indicating that the nificant difference between the conditions (F(3,388) = 3.99,p < .01). mediating pathway is significant. Thus, Study 3 confirmed that the As expected, participants who thought about their own death indi- observed tendency to prefer experiential over material purchases in cated higher WTP for experiential than material purchases (Mmortality response to mortality salience was statistically mediated by differ- = $178.37, SD = 1035.42), compared to those who thought about be- ences in perceived meaningfulness between the experiential and the ing socially excluded (Msocial = -$131.65, SD = 592.77), dental pain material purchases.

(Mphysical = -$154.75, SD = 794.06), and grocery shopping (Mcontrol = -$108.83, SD = 514.34). Follow-up analysis revealed that those Study 4 participants in the social rejection, the dental pain, and the control In Study 4, we tested the effect ofmeaning-fulfillment following condition did not differ on the index of WTP for the experiential and mortality salience threats on consumer preferences for experiential material purchases (p > .97). Thus, Study 2 provided direct evidence over material purchases. We theorized that engaging in an alternative in support of our prediction that reminding people of their own mor- task that allows the individual to recall a past meaningful experience tality leads them to prefer experiential over material purchases and (i.e. meaning-fulfillment) should mitigate the observed tendency for this tendency is only caused by mortality salience (vs. social threats consumers to prefer experiences over material goods in response to vs. physical threats vs. control). mortality salience. Thus, we predicted that the observed tendency will not emerge when people are given an opportunity to fulfill their Study 3 meaning by recalling their past meaningful experience. In Study 3. we explored the underlying mechanism by predicting that people who think about their own mortality (vs. control) will Method and Procedure perceive experiential (vs. material) purchases as more meaningful, This was a 2 (mortality: salient vs. control) x 2 (meaning-fulfill- and the difference in perceived meaningfulness will prompt people ment: yes vs. no) between-participants design. Participants (n = 228) to spend more money on experiences than on material goods in re- were randomly assigned to one of the two conditions. As in Study sponse to mortality salience threats. 2, we manipulated mortality salience (vs. control) by using the writing task and asked participants in the meaning-fulfillment condition Method and Procedure to describe a past experience that they felt was meaningful and to Participants (n = 158) were randomly assigned to either the explain why the event was meaningful to them. Those participants in mortality salience or the control condition. Participants in the mor- the no meaning-fulfillment condition were not asked to describe any tality salience condition were presented a short documentary clip of their past meaningful experiences. Afterward, participants were that reminds them of limited amount of the time left to spend with asked to list three of future experiential and material purchases and their family. In contrast, participants in the control condition were to indicate their WTP for each of purchases they listed. presented a short movie clip about bighorn sheep (Heatherton et al. 1993). Afterward, as in Study 2, participants were asked to list three Results and Discussion of experiential and material purchases and indicate the WTP for each We performed a mortality x meaning-fulfillment ANOVA with of the purchases. They were then asked to rate an item on each of the the index of WTP as a dependent variable. As expected, we found experiential and material purchases that they listed: “To what extent a significant interaction between the mortality salience and the do you think this experiential (material) purchase would be meaning- meaning-fulfillment manipulation (F(1,226) = 4.00, p < .05). In the ful to yourself?” (7-point scale: 1 = “not at all,” 7 = “very much so”). mortality salience condition, participants reported higher WTP for Advances in Consumer Research (Volume 45) / 883 experiential than material purchases when they were not given the Gilovich, Thomas and Amit Kumar (2015), “We’ll Always Have opportunity to recall their past meaningful experience (Mno-fulfillment = Paris: The Hedonic Payoff from Experiential and Material 72.41, SD = 79.66). When they were given the opportunity to recall Investments,” Advances in Experimental Social Psychology, their past meaningful experience, however, the tendency was elimi- 51, 147-187. nated (Mfulfillment= -166.61, SD = 83.14; F(1, 226) = 4.31, p < .05). In Greenberg, Jeff, Tom Pyszczynski, and Sheldon Solomon (1986), the control condition, meaning-fulfillment had no influence on the “The Causes and Consequences of the Need for Self-Esteem: observed effect (Mno-fulfillment = -103.72, SD = 80.99; Mfulfillment= -14.05, A Terror Management Theory,” in Public Self and Private Self, SD = 84.66; F(1, 226) = .59, p > .44). The results of Study 4 dem- ed. Roy F. Baumeister, New York: Springer, 189–212. onstrated recalling one’s past meaningful experience resolved the Greenberg, Jeff, Tom Pyszczynski, Sheldon Solomon, Abram desires for meaning elicited by mortality salience and this, in turn, Rosenblatt, Mitchell Veeder, Shari Kirkland, and Deborah alleviated the tendency to bolster their meaningfulness in response to Lyon (1990), “Evidence for Terror Management II: The Effect mortality salience threats. of Mortality Salience on Reactions to Those Who Threaten or Bolster the Cultural Worldview,” Journal of Personality and General Discussion Social Psychology, 58, 303-318. The current research brings to light one of the key consumption Heatherton, Todd F, Janet Polivy, C. Peter Herman, Roy F. domains in which mortality salience has a substantial influence: Ex- Baumeister (1993), “Self-Awareness, Task Failure, and periential versus material purchases. From a managerial perspective, Disinhibition: How Attentional Focus Affects Eating,”Journal we provide insight for marketing practice, especially for experien- of Personality, 61, 49-61. tial consumption, by emphasizing the need to remind consumers of Kasser, Tim and Kennon M. Shedon (2000), “Of Wealth and Death: the limited amount of time left to them to do everything they ever Materialism, Mortality Salience, and Consumption Behavior,” wanted to do. Thus, it would be key for marketers to understand Psychological Science, 11, 348-351. the meaningful nature of experiences and develop promotion plans King, Laura A, Joshua A. Hicks, Jennifer L. Krull, and Amber (e.g. using images, advertising slogans) which can remind people of K. Del Gaiso (2006), “Positive affect and the experience thoughts about death to differentiate their products from competitors. of meaning in life.” Journal of Personality and Social We hope that the current research spurs other work examining Psychology, 90, 179–196. the nuances of experiential consumption. Further, we would like to Mandel, Naomi and Steven J. Heine (1999), “Terror Management summarize the takeaway from this research with a quote from Bra- and Marketing: He Who Dies with the Most Toys Wins,” zilian novelist Paulo Coelho on his Twitter (@paulocoelho; 2012): Advances in Consumer Research, 26, 527–532. “One day, you will wake up and there won’t be any more time to do Molden, Daniel. C, Gale M. Lucas, Wendi L. Gardner, Kristy Dean, the things you’ve always wanted. Do it now.” and Megan L. Knowles (2009), “Motivations for prevention or promotion following social exclusion: Being rejected versus REFERENCES being ignored,” Journal of Personality and Social Psychology, Arndt, Jamie, Jeff Greenberg, and Alison Cook (2002), “Mortality 96, 415–431. Salience and the Spreading Activation of Worldview-Relevant Richins, Marsha. L. and Scott Dawson (1992), “A consumer Constructs: Exploring the Cognitive Architecture of Terror values orientation for materialism and its measurement: Scale Management,” Journal of Experimental Psychology: General, development and validation,” Journal of Consumer Research, 131, 307-324. 19, 303–316. Arndt, Jamie, Sheldon Solomon, Tim Kasser, and Kennon M. Rosenblatt, Abram, Jeff Greenberg, Sheldon Solomon, Tom Sheldon (2004), “The Urge to Splurge: A Terror Management Pyszczynski and Deborah Lyon (1989), “Evidence for terror Account of Materialism and Consumer Behavior,” Journal of management theory I: The effects of mortality salience on Consumer Psychology, 14 (3), 198–212. reactions to those who violate or uphold cultural values,” Belk, Russel W. (1985), “Materialism: Trait Aspects of Living in Journal of Personality and Social Psychology, 57, 681-690. the Material World,” Journal of Consumer Research, 12, 265- Rosenzweig, Emily and Thomas Gilovich (2012), “Buyer’s 280. Remorse or Missed Opportunity? Differential Regrets for Burke, Brian, Andy Martens, and Erik H. Faucher (2010), “Two Material and Experiential Purchases,” Journal of Personality decades of terror management theory: A meta-analysis of and Social Psychology, 102, 215-23. mortality salience research. Personality and Social Psychology Schlegel, Rebecca J., Joshua A. Hicks, Jamie Arndt, and Laura Review, 14, 155–195 A. King (2009), “Thine Own Self: True Self-Concept Carter, Travis J. and Thomas Gilovich (2010), “The Relative Accessibility and Meaning in Life,” Journal of Personality Relativity of Material and Experiential Purchases,” Journal of and Social Psychology, 96, 473-490. Personality and Social Psychology, 98, 146-159. Steger, Michael F., Patricia Frazier, Shigehiro Oishi, and Matthew _____ (2012), “I Am What I Do, Not What I Have: The Differential Kaler (2006), “The Meaning in Life Questionnaire: Assessing Centrality of Experiential and Material Purchases to the Self,” the Presence of and Search for Meaning in Life,” Journal of Journal of Personality and Social Psychology, 102, 1304- Counseling Psychology, 53, 80-93. 1317. Van Boven, Leaf and Thomas Gilovich (2003), “To Do or to Have? Dunn, Elizabeth and Michael Norton (2013). Happy money: The That Is the Question,” Journal of Personality and Social science of smarter spending. New York: Simon and Schuster. Psychology, 85, 1193-1202. Mind The Gap: How Smaller Numerical Differences Can Increase Product Attractiveness Meyrav Shoham, Bar-Ilan University, Israel Yael Steinhart, Tel-Aviv University, Israel Sarit Moldovan, The Open University of Israel, Israel

EXTENDED ABSTRACT in the version 2.4 vs. 3 condition than when the version number had This research explores situations in which smaller numerical dif- changed from 2 to 3 (F(1,167)=3.60, p=.06). It was also more positive ferences can be perceptually larger and consequently increase product than in the version 2 vs. 2.6 condition (F(1,167)=4.98, p=.03), and attractiveness. We find that when a product’s numerical information the version 2.1 vs. 2.7 condition (F(1,167)=8.75, p<.01). Differences changes from a decimal to an integer (e.g., 3.4 to 4), consumers are between the other conditions in the study were not significant. These likely to perceive the product as more improved and therefore more ap- results indicate that it is not the precision of a decimal number in itself pealing than when the change is between two integers (e.g., from 3 to that is advantageous, but rather the combination of numerical precision 4), even though the difference is larger in the latter case. Specifically, and the crossing of a numerical threshold. we propose that when a product’s version number or rating changes Study 3 provided support for the underlying mechanism of per- from a decimal number to an integer, consumers infer that the product ceived improvement. Participants (n=85) were randomly assigned to has leapt over intermediate values and crossed the threshold into a new one of two conditions: a hotel whose previous “eco-friendly rating” category, reflecting greater improvement. was either 8 (integer condition) or 8.3 (decimal condition). They read A decimal number can draw attention to the fine-grained scale that this rating had increased to 9 following a recent renovation. We and may suggest the possibility of additional, intermediate values found that the hotel was perceived as better when its rating had in- (Pandalaere, Briers and Lermbregts 2011). Consequently, we propose creased from 8.3 to 9 compared to when it had increased from 8 to 9 that the change from a decimal to an integer (rather than to another (t(83)=2.35, p=.02). Interest in staying at the hotel was also greater in decimal number) is perceived as the crossing of a threshold. Building the decimal-to-integer rating condition compared to the integer-to-in- on research showing the impact of category boundaries on consumers’ teger rating condition (t(83)=1.92, p=.06). Perceived improvement of perceptions and behavior (e.g., Irmak, Walker-Naylor, and Bearden the hotel mediated the effect (b=0.20, SE=0.11; 95% CI: 0.04 to 0.46). 2011; Isaac and Schindler 2013), we predict that a decimal-to-integer In study 4, we examined a possible moderator: the type of numer- boundary crossing will lead consumers to infer that the product has ical information that consumers encounter. Participants (n=190) were improved substantively, which will increase product attractiveness. randomly assigned to one of four conditions in a 2(previous product This positive effect of a smaller, more precise difference is con- model: decimal or integer) × 2(type of numerical information: familiar tingent upon the perception that a category threshold has been crossed. or less familiar) experimental design. They were asked to imagine that Therefore, we do not expect it to occur when the change is from one they wanted to buy a new camera. In the familiar condition, partici- decimal number to another (e.g., from 2.1 to 2.7), as this is likely to be pants were told that the previous model of this camera had a display perceived as a relatively minor improvement. If the integer component size of 3 or 3.4 inches, while the new camera had a 4-inch display. of a number remains unchanged, consumers may focus on that rather In the less familiar condition, participants were told that the camera’s than on changes in the digits after the decimal point. “color accuracy rating” had increased from 3 or 3.4 to 4. We found a fa- The decimal-to-integer effect also depends on consumers’ familiarity × type of numerical information interaction (F(1,186)=5.98, miliarity with the product and the scale. When the scale is vague or p=.02), such that participants who were presented with the unfamiliar unfamiliar, consumers need to infer the meaning of the numerical color accuracy rating were more interested in buying the camera when information from whatever cues are available which will lead to the the rating had improved from 3.4 to 4 than when it had increased from decimal-to-integer effect. If consumers have existing knowledge or 3 to 4 F(1,186)=4.13, p=.04). Among participants who read about the understanding as to what the number actually represents, they should camera’s display size, the difference between the decimal and integer be less prone to such inferences about the meaning of the numerical conditions did not reach significance (F(1,186)=2.05, n.s.). differences. While larger numerical gaps might be expected to signal a greater In four studies, we provide support for the proposed effect, and difference between product versions or ratings, our findings indicate demonstrate the mediating role of product improvement and how (lack that sometimes the reverse is true: smaller differences can be percep- of) familiarity with the numerical information moderates the effect. tually larger and boost product attractiveness. In other words, some Study 1 demonstrated that a smaller difference between a decimal small differences feel larger than others – and it is important to “mind version number and an integer version number can increase perceived the gap” and its implications. product attractiveness. We randomly assigned undergraduate participants (n=96) to one of two conditions: software whose existing version REFERENCES number was either 3 (integer condition) or 3.4 (decimal condition). Irmak, Caglar, Rebecca Walker Naylor, and William O. Bearden. The new version in both conditions was 4. We found that participants (2011), “The out-of-region bias: Distance estimations based on who were told that the upgrade under consideration was from version geographic category membership,” Marketing Letters, 22 (2), 3.4 to version 4 were more favorable about an upgrade than those who 181-196. were told that the upgrade being considered was from version 3 to ver- Isaac, Mathew S., and Robert M. Schindler (2013), “The Top-Ten sion 4 (t(94)=2.16, p=.03). Effect: Consumer’s Subjective Categorizations of Ranked In study 2, participants (n=172) were randomly assigned to one Lists,” Journal of Consumer Research, 40 (2), 1181–1202. of five conditions in which they read descriptions of two versions of Pandelaere, Mario, Barbara Briers, and Christophe Lembregts a “social camera,” with different combinations of version numbers. A (2011), “How to make a 29% increase look bigger: The unit one-way ANOVA revealed significant differences between the condi- effect in option comparisons,”Journal of Consumer Research, tions (F(4,167)=2.57, p=.04). Camera evaluations were more positive 38 (2), 308-322.

Advances in Consumer Research 884 Volume 45, ©2017 The Friend-Number Paradox Kao Si, The Chinese University of Hong Kong, Hong Kong Xianchi Dai, The Chinese University of Hong Kong, Hong Kong Robert Wyer, The Chinese University of Hong Kong, Hong Kong

EXTENDED ABSTRACT with the individual. We propose that this ability is negatively affected Little research has examined people’s competence in predicting by an individual’s number of social ties because limited resources others’ perceptions of their own attractiveness in social networks. become more diluted with a larger number of social ties. Therefore, Mispredictions of this kind are consequential: They hinder the estab- a person would expect the quality of his or her relationship with a lishment and maintenance of desirable relationships and negatively given individual to be poorer if that individual had a large number of affect the efficient allocation of resources to achieve particular goals friends, and be less likely to make the effort to become friends with in the social realm. For example, a woman might overestimate oth- that individual. In other words, we predict that people are more likely ers’ liking for her and adopt an interaction strategy that actually de- to network with individuals who have fewer existing friends. creases her popularity (Dai, Dong, and Jia 2014). A political candi- Preference Mispredictions in Social Networking date who over-predicts his or her favorability in the eyes of minority If the preceding observations are valid, then why would people groups might distribute his or her attention and resources in a subop- overlook others’ concern for relationship quality and mispredict their timal manner that decreases the chances of winning. In short, correct preferences? For one thing, taking the perspective of others requires insights into others’ preferences towards oneself are important for effort and motivation (Epley, Keysar, Van Boven, and Gilovich 2004; people’s well-being and functioning in general and for accomplish- Lin, Keysar, and Epley 2010). As a result, people often exhibit an ment of effective marketing campaigns in particular. egocentric bias, focusing on self-relevant information while paying In this research, we propose and investigate a fundamental insufficient attention to concerns that might underlie others’ judg- preference mismatch in social networking with regard to individu- ments (Ross and Sicoly 1979; Windschitl, Kruger, and Simms 2003). als’ numbers of friends. By “social networking,” we are referring It is therefore proposed that when predicting the number of social ties to the will and practices undertaken to expand one’s social circle. It that others prefer, people are prone to overlooking others’ concern includes but is not limited to the use of Internet-based social media for relationship quality and focus instead on the symbolic signifi- platforms. In addition, we focus on social networking for the pur- cance of social ties. In other words, they may infer that the greater pose of establishing “pure” friendships rather than those formed in the number of ties they have, the higher their social esteem and more pursuit of economic, professional, or other interests (e.g., establish- likeable their personality traits, and thus the stronger others’ prefer- ing business connections). We show that people expect others to be ence for networking with them. more willing to make friends with them when they already have a Furthermore, we propose that this preference mismatch is also large number of friends, whereas others actually exhibit a preference a result of people’s neglect of their own obligations and expecta- that is diametrically opposed to that expectation: People are actually tions in the social network when formulating the predictions. Re- more willing to make friends with someone who has a small number search shows that people are inclined to neglect various kinds of “ex- of friends. We coin the term friend-number paradox to describe this penses” when they make judgments and predictions (Berman, Tran, mismatch between people’s expectation and others’ actual prefer- Lynch and Zauberman 2016; Campbell and Warren 2015; Frederick, ence. In the following, we explicate the reasons behind this paradox Novemsky, Wang, Dhar, and Nowlis 2009). For example, they often and report the results of four studies confirming the validity of our neglect opportunity costs in their decisions and choices (Frederick et propositions. al. 2009). If insufficient consideration has been given to the obligations and expectations inherent in social ties, individuals are likely to Theoretical Background become insensitive to the decrease in relationship quality caused by Preference for Small Number of Friends increases in the numbers of social ties, and thus express a preference Social ties are the building blocks of the social world and are for a larger number of ties in their predictions. thus highly valued, with larger numbers of social ties often reflecting In summary, we propose that people expect others to be more higher social esteem and more likeable personality traits. It there- willing to make friends with them when they have a large number fore seems reasonable to assume that people prefer individuals with of friends. Others, in contrast, are more willing to make friends with larger numbers of friends as social networking targets (Feld 1991). someone who has a small number of friends. This preference mis- However, social ties are valuable only to the extent that adequate match arises because people overlook others’ concern with relation- levels of relationship quality are maintained, and relationship quality ship quality and focus on the symbolic meaning of social ties when is essentially entrenched in the obligations and expectations inherent making predictions about others. Further, the difference in relation- in social ties. Reliability and the due fulfillment of obligations and ship quality concern is closely related to neglect of one’s obligations expectations cost attention, time, and other resources. to and expectations from friends at the time these predictions are In his seminal paper, Coleman (1988) suggested that obligations formulated. We conducted four studies that demonstrate this prefer- and expectations constitute a major form of social capital that facili- ence mismatch and test the validity of our hypotheses. We describe tates actions within social networks and is an important resource with the major findings of the studies in the next section. regard to efficiency and well-being. Given the importance of social capital, its accumulation is an implicit yet fundamental objective in Study 1: Friend Number on Facebook social networking. In other words, relationship quality is the primary Method concern of social networking. Hence, an individual’s ability to honor 102 participants (M = 35.21, 59% male) took part in a short his or her obligations to another person and to meet that person’s ex- Age choice study. The participants read that Facebook is one of today’s pectations (i.e., ability to build social capital with that person) should most widely used online social networking sites. Then, those in the be a predominant predictor of the latter’s preference for networking

Advances in Consumer Research 885 Volume 45, ©2017 886 / The Friend-Number Paradox others’ preference condition were asked to predict whether others friends with a person with many good friends over one with an equal would be more likely to make friends with them if their Facebook number of acquaintances. The difference between the two conditions page showed that they had 500 friends or 50 friends. Participants was highly significant χ( 2(1) = 12.51, p < .001). These results there- in the own preference condition were asked whether they would be fore provide further evidence that the preference mismatch is driven more likely to make friends with a person whose Facebook page by concerns over relationship quality rather than inferences about the showed 500 friends or 50 friends. target person’s social esteem or personality traits.

Results Study 3: Concern With Relationship Quality 71% (36/51) of participants in the others’ preference condition Study 3 provided evidence supporting the mediating role of dif- predicted that others would be more likely to make friends with them ferences in relationship quality concern in the current effect. if they had 500 rather than 50 friends on Facebook. In contrast, only Method 31% (16/51) of participants in the own preference condition pre- 100 participants (M = 38.89, 52% male) took part in this ferred an individual with 500 rather than 50 Facebook friends. The Age 2 study. Study 3 adopted a design similar to that of Study 1. After difference between the two conditions was highly significantχ ( (1) participants made their choices, they were given four descriptions = 15.69, p < .001). regarding their concerns for relationship quality and were asked to indicate whether those concerns had occurred to them when they Study 2: Preference for a Target Person’s Relationship were making their choices. Quality The results of Study 1 are consistent with our proposed effect. Results However, an alternative explanation could also account for our find- The choice data replicated results of our previous studies. 72% ings: Individuals with an excessive number of friends are associated (36/50) of the participants in the others’ preference condition pre- with certain negative impressions (e.g., frivolousness) that affect dicted that another individual would be more likely to make friends others’ inferences about their social esteem and personality traits. with them when they had 200 instead of 50 friends. However, only Therefore, participants in the own preference condition may have 22% (11/50) of participants in the own preference condition ex- chosen the person with fewer friends not because of relationship pressed a preference for making friends with another individual with quality concerns but out of concerns over the target person’s social 200 versus 50 friends (χ2(1) = 25.09, p < .001). esteem or personality traits. Analysis of the combined measure of relationship quality con- In Study 2, we tested this alternative explanation by asking par- cern (α = .90) revealed that the participants in the own preference ticipants whether they would prefer a target person who had many condition were concerned with relationship quality to a significantly good relationships (strong relationship quality) versus another who greater extent (M = 5.20, SD = 1.50) than those in the others’ pref- had merely many acquaintances (weak relationship quality). Like erence condition (M = 4.33, SD = 1.44; F(1, 98) = 8.88, p = .004; relationship quantity, average relationship quality should also nega- d = .59). We confirmed the mediating effect of relationship qual- tively influence an individual’s ability to build social capital with ity concern on the relationship between judgment perspective and others. However, stronger average relationship quality should be un- friend-number preference using the bootstrapping method (Hayes conditionally associated with higher social esteem and more likeable 2013). With 5000 bootstrap samples, the indirect mediating effect personality traits. If the effects in Study 1 were indeed driven by was estimated to be -.93 (boot SE = .57), with its 95% bias-corrected differing concerns over relationship quality, then we would expect confidence interval (-2.27, -0.23) excluding zero. a significant difference in preferences to again emerge between the two judgment conditions. Study 4: Reminder of Obligations and Expectations As previously noted, differences in the degree of concern with Method relationship quality are inextricably associated with the failure to 101 participants (M = 35.61, 62% male) took part in this Age consider the obligations and expectations inherent in social ties when study. Participants in the others’ preference condition read a scenario predicting the preferences of others. Hence, reminding participants in which they were asked to imagine themselves in each of two situa- of those obligations and expectations should attenuate or even re- tions. In one situation, they had lived in a community for a long time verse a preference for large numbers of friends in their predictions, and had good relationships with many people. In the other, they had thereby bringing those predictions more in line with the actual pref- recently moved to a new community and had made the acquaintance erences of others. Study 4 examined this prediction. of the same number of people as in the first situation. They then selected the situation in which they thought others would be more Method willing to make friends with them. Study 4 employed a similar design as in Studies 1 and 3. A In the own preference condition, participants were instructed to reminder condition was added in which participants were asked to imagine two fictitious persons in their community who were similar think about and write down what others would expect them to do in to each other in all aspects except that one had lived in the commu- various situations as friends before they predicted others’ preference nity for a long time and had good relationships with many people, for their number of friends in social networking. 205 participants whereas the other had moved to the community recently and was (MAge = 33.31, 59% male) were randomly assigned to three between- merely acquainted with the same number of people. They were then participants conditions. asked to select the person with whom they would prefer a friendship. Results Results The participants’ choices differed significantly among the three Of the participants in the others’ preference condition, 84% conditions (χ2(2) = 31.18, p < .001). Replicating our previous find- (42/50) thought that others would be more likely to make friends ings, the majority of participants (79%; 56/71) in the own prefer- with them if they had many good friends rather than an equal num- ence condition preferred another individual with 50 rather than ber of acquaintances. In the own preference condition, however, 200 friends, whereas only a small proportion of participants (32%; only 51% (26/51) of participants expressed a preference for making 23/71) in the others’ preference (baseline) condition correctly pre- Advances in Consumer Research (Volume 45) / 887 dicted that preference by the majority of others (χ2(1) = 31.07, p Coleman, James S. (1988), “Social capital in the creation of human < .001). More importantly, compared with the baseline condition, a capital,” American Journal of Sociology, 94, S95-S120. significantly larger percentage (57%; 36/63; χ2(1) = 8.30, p = .004) Dai, Xianchi, Ping Dong, and Jayson S. Jia (2014), “When does of participants in the reminder condition predicted that others would playing hard to get increase romantic attraction?” Journal of be more likely to make friends with them if they had 50 rather than Experimental Psychology: General, 143 (2), 521-6. 200 friends, although the percentage still differed from that in the Epley, Nicholas, Boaz Keysar, Leaf Van Boven, and Thomas own preference condition (χ2(1) = 7.33, p = .007). These findings Gilovich (2004), “Perspective taking as egocentric anchoring again provide support for our proposed mechanism underlying the and adjustment,” Journal of Personality and Social friend-number paradox. Psychology, 87 (3), 327-39. Feld, Scott L. (1991), “Why your friends have more friends than General Discussion you do,” American Journal of Sociology, 96 (6), 1464-77. One important motivation for engaging in social networking is Frederick, Shane, Nathan Novemsky, Jing Wang, Ravi Dhar, and to expand one’s number of social ties. In the studies reported herein, Stephen Nowlis (2009), “Opportunity cost neglect,” Journal of we found that in the process of such engagement, people paradoxi- Consumer Research, 36 (4), 553-61. cally prefer individuals with a relatively small rather than large num- Hayes, Andrew F (2013), Introduction to mediation, moderation, ber of friends. At the same time, they are rather inept at predicting and conditional process analysis: A regression-based that preference of others, judging instead that others would prefer to approach, Guilford Press. network with them if they had more rather than fewer friends. Lin, Shuhong, Boaz Keysar, and Nicholas Epley (2010), The current research examines a fundamental mismatch be- “Reflexively mindblind: Using theory of mind to interpret tween people’s predictions of others and their own personal pref- behavior requires effortful attention,” Journal of Experimental erences. We investigated the mismatch in the context of social Social Psychology, 46 (3), 551-6. networking, which is highly relevant to daily life and important to Ross, Michael, and Fiore Sicoly (1979), “Egocentric biases in well-being. However, we believe that the friend-number paradox is availability and attribution,” Journal of Personality and Social pertinent to a much wider scope of contexts, and hope that further Psychology, 37 (3), 322-36. studies will be inspired to explore its broader implications. Windschitl, Paul D., Justin Kruger, and Ericka Nus Simms (2003), “The influence of egocentrism and focalism on REFERENCES people’s optimism in competitions: When what affects us Berman, Jonathan Z., An TK Tran, John G. Lynch Jr, and Gal equally affects me more,” Journal of Personality and Social Zauberman (2016), “Expense Neglect in Forecasting Personal Psychology, 85 (3), 389-408. Finances,” Journal of Marketing Research, 53 (4), 535-50. Campbell, Margaret C., and Caleb Warren (2015), “The progress bias in goal pursuit: When one step forward seems larger than one step back,” Journal of Consumer Research, 41 (5), 1316- 31. Time Units and Patience Rafay A. Siddiqui, Hong Kong Polytechnic University, Hong Kong Ashwani Monga, Rutgers University, USA Eva Buechel, University of South Carolina, USA

EXTENDED ABSTRACT In study 2, participants were first manipulated to think of money Consider a consumer in the market for an iPad, who could pur- as a means to buy something hedonic versus utilitarian. Then they chase the current generation iPad (SS) or wait for the next generation were offered either a small monetary reward that they could receive version that is better, but available later in time (LL). In the current immediately or a larger monetary reward that they could receive lat- work, we show that patience may increase when wait time for LL is er. The wait time for the larger reward was expressed as 4 days vs. 96 expressed using larger units of time. Importantly, we argue that this hours. Participants waited for the larger amount more when wait time only occurs when the rewards are hedonic in nature, and not when was expressed in days rather than hours, but only when they thought they are utilitarian. of money as a means to buy something hedonic (not utilitarian). In study 3, participants imagined receiving a $100 visa gift card Theoretical Development and were asked to think about what they will spend it on. They were Patience in intertemporal choices may be affected by wait time then informed that they could receive a $110 gift card instead if they perception (Kim, Zauberrman and Bettman 2012), as well as the na- were to wait. Wait time was expressed as 14 days vs. 2 weeks. Partic- ture of the rewards (Thaler 1981). We demonstrate a novel interac- ipants preferred to wait more when wait time was expressed in weeks tive effect of wait time and rewards. Specifically, we show that the rather than days, but only when they thought about buying hedonic units used to describe wait time (e.g., days vs. months) influence wait (not utilitarian) products with the gift card. Participants’ wait time time perception, and consequently patience, but this is contingent on perception mediated the effect of units and reward type on patience. whether the rewards are hedonic or utilitarian. Our prediction for this In study 4, participants imagined buying headphones online, interactive effect is based on past work on the numerosity heuristic which were manipulated to be either hedonic or utilitarian. They and the hedonic vs. utilitarian distinction in consumer products. chose between standard and expedited shipping, with shipping times People have a tendency to equate smaller numbers with smaller expressed in hours vs. days. For hedonic (but not utilitarian) head- magnitudes, an effect known as the numerosity heuristic (Pelham, phones, participants chose standard shipping more, and thus reflected Sumarta and Myaskovsky 1994). For example, 10,000 Korean Won more patience, when shipping times were in days rather than hours. appear to be greater than 1,000 Japanese Yen, even though both are Participants’ wait time perception mediated the effect of units and of equal value (Raghubir and Srivastava 2002). Thus, if the wait time reward type on patience. for LL is expressed as 3 months instead of 90 days, people should In study 5, participants were first manipulated to think of money perceive it to be shorter and, consequently, become more patient. The as a means to buy something hedonic versus utilitarian. They then in- reason for this is that the larger unit of months results in a smaller dicated preference between a $100 gift card in 2 weeks (vs. 14 days) number being used to express the wait time (i.e., ‘3’ rather than ‘90’). versus a $120 gift card in 8 weeks (vs. 56 days), and completed the However, since numerosity is an effortless heuristic, it should impact lay rationalism scale (Hsee et al. 2015) as a measure of their natural intertemporal choice more when decision makers are relying less inclination towards relying more on calculation. Results from earlier on careful calculation. We believe one such situation is when indi- studies replicated, but only for participants who were relatively low viduals evaluate hedonic rather than utilitarian rewards. Given that on the lay rationalism scale. the primary purpose of hedonic consumption is pleasure rather than In study 6, in a control condition, participants completed the practicality (Khan et al. 2005; Pham 1998), hedonic consumption same procedure as in study 4. In a calculation prime condition, paris associated less with careful thinking and more with reliance on ticipants first solved a few mathematical problems that would en- feelings. The influence of the numerosity heuristic in intertemporal courage careful calculation rather than the use of heuristics. Results choice should thus be stronger in the case of hedonic rewards. Hence, from study 4 replicated in the control condition, but in the calculation we predict that larger time units should lead to shorter wait time per- prime condition no significant effects were observed even when the ception and, consequently, higher patience, but that this should only rewards were hedonic. be the case when the rewards are hedonic rather than utilitarian. We observe support for our prediction in six studies, four of Discussion which provide process evidence. Studies 3 to 6 show that the interac- We show that using larger units to express wait time can boost tive effect of time unit and reward type on patience is mediated by the patience, but only when the rewards are hedonic and not when they perceived length of the wait time. Studies 5 and 6 demonstrate that are utilitarian. Reliance on careful calculation attenuates the effect this interactive effect attenuates when participants engage in careful even for hedonic rewards. These results reveal a novel interactive calculation. effect between the nature of the rewards and wait time perception in intertemporal choice. Studies In study 1, participants made a hypothetical choice between the REFERENCES current version of a computer tablet versus an upcoming version. Hsee, Yang Yang, Xingshan Zheng, and Hanwei Wang (2015), “Lay The tablet was manipulated to be either hedonic or utilitarian and the Rationalism: Individual Differences in Using Reason Versus wait time for the upcoming version was expressed as 120 days vs. Feelings to Guide Decisions,” Journal of Marketing Research, 4 months. Participants were more patient for the new version when 52 (1), 134-146. wait time was expressed in months rather than days, but only when the tablet was hedonic (not when it was utilitarian).

Khan, Uzma, Ravi Dhar, and Klaus Wertenbroch (2005), “A Pham, Michel T. (1998), “Representativeness, Relevance, and the Behavioral Decision Theory Perspective on Hedonic and Use of Feelings in Decision Making,” Journal of Consumer Utilitarian Choice,” Inside Consumption: Frontiers of Research, 25 (2), 144-59. Research on Consumer Motives, Goals, and Desires, 144-65. Raghubir, Priya, and Joydeep Srivastava (2002), “Effect of Face Kim, B. Kyu, Gal Zauberman, and James R. Bettman (2012), Value on Product Valuation in Foreign Currencies,” Journal of “Space, Time, and Intertemporal Preferences,” Journal of Consumer Research, 29 (3), 335-47. Consumer Research, 39 (4), 867-80. Thaler, Richard H. (1981), “Some Empirical Evidence on Dynamic Pelham, Brett W., Tin T. Sumarta, and Laura Myaskovsky (1994), Inconsistency,” Economics Letters, 8 (3), 201-07. “The Easy Path from Many to Much: The Numerosity Heuristic,” Cognitive Psychology, 26 (2), 103-33. Lacking a Resource or Being One: Money Scarcity versus Time Scarcity Differentially Shape Self-value and Product Judgments Jane So, University of Washington, USA Nidhi Agrawal, University of Washington, USA

EXTENDED ABSTRACT resource). Moreover, we found that the increased positive self-in- In this article, we investigate how time scarcity and money ference significantly predicted self-value (e.g., self importance, self- scarcity differentially affect consumer’s subjective assessment of the esteem) and self-value significantly predicted consumer’s preference self-value via altering consumer’s view of the self as a resource. towards self-value promoting advertisement. We found a serial me- Time and money are both primary resources that are used to diation in this study. achieve certain ends (Becker 1965). However, consumer research- In study 2, we showed the effect of time (vs. money) scarcity on ers have shown that time is systematically different from money in positive self-inference can be switched off by using personal control multiple ways (Leclerc et al. 1995; Mogilner and Aaker 2009; Okada as a moderator. If time scarcity leads to greater positive self-infer- and Hoch 2004). One important difference between time and money ence because spending time requires the self as an actor, then those is that money does not require the self as an actor to have its value who perceives themselves as unable to change the outcome will be and meaning but time is bounded up with the self to have its value less likely to infer that they are the scarce resource. As predicted, we and meaning. Money is a medium of exchange that has a fixed value replicated our previous effect that time (vs. money) scarcity leads to (Leclerc et al. 1995; Okada and Hoch 2004; Saini and Monga 2008). greater positive self-inference, among those who have high personal However, time is by its definition, an experience that the self is con- control. However, we found that such difference was diminished nected to (Liu and Aaker 2008; Mead 1934; Reed et al. 2007). among those who have low personal control. We suggest that due to this difference in interconnectedness In study 3, we showed that money scarcity effect can be switched of the self and the resources, the self-inferences that people make on. If money scarcity does not lead to positive self-inference because when feeling time (vs. money) scarcity can be different. Because the self as an actor is not salient to consumers, then highlighting the money is not bounded up with the self, the self-inference that people positive role of the self should boost their positive self-inference. make when feeling money scarcity is simply that ‘I don’t have the We tested this using the self as the source of scarcity. Those who resource’. Money is a tangible resource that one can possess (Mac- view themselves as a positive source of scarcity (i.e. feeling scar- donnell and White 2015) and be used to exchange with desirable city because the self was prioritizing the use of resource), we should outcomes. Thus, people who are feeling money scarcity may have a see that even money scarcity leads to greater positive self-inference. negative self-inference such that they lack the scarce and important However, those who view themselves as a negative source of scar- resource. However, time is a resource that the self is at root (Flaherty city (i.e., feeling scarcity because the self mismanaged the use of and Fine 2002; Mead 1932), thus scarcity of time entails meanings resource), we should replicate our previous effect. We used 2(money other than simply not possessing the resource. Because individual’s scarcity vs. time scarcity) x 2(prioritization vs. mismanagement vs. activities and experiences translates into spending time (George and control) between subjects design. We replicated our previous find- Jones 2000; Mead 1932), lack of time means the shortage of the self ings in the control and the mismanagement condition, such that time as an actor who needs to take multiple roles at the same time (Jabs scarcity condition indicated greater positive self-inference and pref- and Devine 2006; Kaufman et al. 1991). Thus when feeling time erence towards indulgent product than the money scarcity condition. scarcity, people may have a positive self-inference such that ‘I am the However, within the prioritization condition, the money scarcity con- scarce resource’, over and above the negative inference of ‘I don’t dition showed increased positive self-inference and increased prefer- have the scarce resource’. ence towards indulgent product. We further suggest that this relative difference in self-inferences Our findings suggest that compared to money scarcity, time will differentially affect consumer’s self-value. We suggest that the scarcity leads to positive self-inference whereas both time scarcity desirability and positive aspects of scarcity can be applied to the self and money scarcity lead to negative self-inference. This is due to that when making the positive self-inference that further leads to greater time is more bounded up with the self, such that spending time can- self-value. We measure the self-value through a variety of self-relat- not be done without the self as an actor. Such positive self-inference ed variables such as self-esteem, self-importance, and self-achieve- leads time (vs. money) scarce people to have greater assessment of ment. Such increased self-value shapes consumer’s preference to- self-value. wards products that reflect the higher self-value such as choosing higher priced products and better evaluations of products that pitch REFERENCES the self as ‘being worth it’ or special. Furthermore, to test that these Becker, Gary S. (1965), “A Theory of the Allocation of Time,” effects are driven by the positive self-inferences, we document con- Economic Journal, 75 (299), 493–517. ditions when time scarcity does not lead to favorable self-inferences Flaherty, Michael G., and Gary Alan Fine (2001), “Present, Past, (i.e., low personal control) and when money scarcity leads to more and Future,” Time & Society 10 (2-3), 147-161. favorable self-inferences (e.g., when the scarcity arises from priori- George, Jennifer M., and Gareth R. Jones (2000), “The Role tization). of Time in Theory and Theory Building,” Journal of The objective of study 1 is to show that time (vs. money) scar- Management, 26(4), 657-684. city leads to increased positive self-inference, which further affects Jabs, Jennifer, and Carol M. Devine (2006), “Time Scarcity and consumer’s responses. We used pre/post measure of positive and Food Choices: an overview,” Appetite, 47 (2), 196-204. negative self-inferences to test this link. We found that both time Kaufman, Carol F., Paul M. Lane, and Jay D. Lindquist (1991), and money scarcity led to greater negative self-inference (i.e., I lack “Exploring More than 24 Hours Day: A Preliminary the scarce resource). However, we found that only time (vs. money) Investigation of Polychronic Time Use,” Journal of Consumer scarcity led to greater positive self-inference (i.e. I am the scarce Research, 18 (3), 392–401.

Leclerc, France, Bernd H. Schmitt, and Laurette Dube (1995), Mogilner, Cassie, and Jennifer Aaker (2009), ““The Time vs. “Waiting Time and Decision Making: Is Time like Money?” Money Effect”: Shifting Product Attitudes and Decisions Journal of Consumer Research, 22 (1), 110-19. through Personal Connection,” Journal of Consumer Research, Liu, Wendy and Jennifer Aaker (2008), “The Happiness of Giving: 36 (2), 277-91. The Time-Ask Effect,” Journal of Consumer Research, 35 (3), Okada, Erica Mina, and Stephen J. Hoch (2004), “Spending Time 543–57. versus Spending Money,” Journal of Consumer Research, Macdonnell, Rhiannon, and Katherine White (2015), “How 31(2), 313-23. Construals of Money Versus Time Impact Consumer Reed, Americus II, Karl Aquino, and Eric Levy (2007), “Moral Charitable Giving,” Journal of Consumer Research, 42 (4), Identity and Judgments of Charitable Behaviors,” Journal of 551-63. Marketing, 71 (1), 178–93. Mead, G. H. (1932), The Philosophy of the Present, Chicago, Saini, Ritesh, and Ashwani Monga (2008), “How I Decide Depends London: Open Court Publishing. on What I Spend: Use of Heuristics Is Greater for Time than for Money,” Journal of Consumer Research, 34 (6), 914–22. The Power of Negative Visualizations: When Fear Appeals Motivate Consumer Behavior Kamila Sobol, Concordia University, Canada Marilyn Giroux, Auckland University of Technology, New Zealand

EXTENDED ABSTRACT their daily routine in a general sense (i.e. abstract) or specifically Fear appeals are a common marketing tactic that intend to per- visualized their morning routine (i.e. concrete). In a seemingly un- suade consumers to buy products and services that aim to help avoid related study participants were asked to write down their new year’s an undesirable outcome. It is proposed that the main mechanism in- resolutions. Answers were coded for how many unrelated goals were volved is the arousal of fear or anxiety, a potent source of energy that reported (dependent variable). If the avoidance goal (i.e. feared body propels people into action (Hull 1943). It is assumed that scare tactics appearance) was still activated, participants should engage in goal should modify behavior when the undesirable outcome is perceived shielding and list less goals unrelated to health and fitness. An ANO- as severe, probable, yet surmountable (Rogers 1975). However, em- VA and simple contrasts verified that abstract fear visualizations pirical evidence reveals that fear appeals are not always efficient motivated goal shielding tendencies more so than the other types despite meeting these criteria (e.g. Morales et al. 2012; Passyn and of visualizations (Mviz_abstract=.85, Mviz_concrete=1.57, Mcontrol_abstract=1.79,

Sujan 2006). Various variables were examined to explain the incon- Mcontrol_concrete=1.33). sistencies in past findings (e.g. Manyiwa and Brennan 2012; Smith and Stutts 2003), focusing primarily on moderators relating to types Study 2 of fear appeal, emotions and individual differences. The main objective of study 2 was to confirm that abstract vi- The present paper employs the technique of visualization to fur- sualizations of the feared self motivate goal-congruent choices. 194 ther understand instances when fear appeals are, and are not, efficient undergraduate students participated in a 2 (visualization task: feared at motivating behavior. Work on negative visualizations is limited self, control) x 2 (processing style: abstract, concrete) between-sub- and generally affirms that mental imagery of feared self-states en- ject design. 12 participants (6%) were dropped for failing to com- courage consumers to engage in behaviors that will distance them plete the imagery task. Participants first completed the same visu- from the undesirable outcome (Hoyle and Sherrill 2006). However, alization task as in study 1. After a short filler task, subjects were we propose that negative visualizations are not all equally efficient, presented with a list of 20 consumer goods, and rated how likely they and that processing style (abstract versus concrete) is a key determi- were to purchase each item (7-point scale). Included in the list were nant of the fear appeal’s persuasiveness. More specifically, abstract 6 health-related items (e.g. running shoes, smoothie maker). The imagery of the feared self should increase motivation to avoid the mean score served as the dependent variable. An ANOVA and simple undesirable outcome, while concrete visualizations do not produce contrasts verified that abstract visualizations of the feared self moti- such motivational tendencies. vated higher intent to purchase healthy goods, compared to the other

These propositions are based on past research which affirms that visualizations (Mviz_abstract=5.12, Mviz_concrete=4.55, Mcontrol_abstract=4.65, abstract processing reduces perceived boundaries between an indi- Mcontrol_concrete=4.61). vidual’s various self-concepts, while concrete processing heightens self-concept distinctions (Forster et al. 2008; Wakslak et al. 2008). Study 3 This theorizing suggests that mentally simulating one’s feared self- The objective of Study 3 was to provide evidence for the pro- state in abstract terms should evoke perceptions of assimilation to posed affect-based mechanism, namely that abstract imagery evokes the imagined feared self, which is likely to evoke sufficient anxiety feelings of anxiety, and that these feelings prompt motivated be- required to drive motivated behavior. Conversely, visualizations pro- havior. 215 participants were randomly assigned to the abstract fear cessed in a concrete fashion heighten perceptions of self-discrepan- and routine visualization conditions (the concrete conditions were cy, thereby reassuring individuals that the feared self is distant from removed). They then reported their feelings, and completed the same their present state and hence, that no action is required. In sum, we consumer survey as in the previous study, measuring their level of in- propose that fear appeals have the potential to motivate consumer terest in health-related products. A mediational analysis using model behavior, but only when they instigate abstract (not concrete) visu- 4 of Hayes’s PROCESS macro indicated a partial mediation (5000 alizations. Further, we also show that this effect is driven by feelings samples, 95% CI,.04-.13) (figure 1). The indirect effect of abstract of anxiety. fear visualizations on healthy behaviors through anxiety was positive, however, anxiety only partially mediates this relationship since Study 1 the direct effect of abstract fear visualizations on healthy behaviors, To verify whether abstract visualizations enhance motivation, after the path through anxiety was accounted for, was still significant. we examine whether they increase people’s tendency to engage in goal shielding – i.e. a tendency to inhibit other goals from diverting General Discussion attention from a focal goal (Shah et al. 2002). 128 participants were The present findings show when and how visualizations of the randomly assigned to a 2 (visualization task: feared self, control) by feared self motivate people to make goal-relevant consumer choices 2 (processing style: abstract, concrete) between-subject design. 12 and engage in goal-congruent behaviors, as well as provide prelimi- participants (9%) were dropped for failing to complete the manipula- nary evidence for the mediating role of affect in this relationship. The tion task. In the feared self condition, participants either visualized present paper contributes to the literature in at least three ways. First, having achieved their feared body appearance in a general sense (i.e. we are the first to employ the visualization technique when examin- abstract processing), or first specified important attributes associ- ing the effects of fear appeals in consumer research. Second, we are ated with their feared body image (e.g. gain 20 pounds, cellulite) the first to examine the role of processing style as a moderator in and then visualized having achieved these concrete attributes (i.e. the mental imagery literature. Third, we provide empirical evidence concrete processing). In the control condition, participants visualized

Advances in Consumer Research 892 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 893 that negative mental imagery motivates consumer behavior via emo- Ruiter, R., Abraham, C., & Kok, G. (2001). Scary warnings and tional arousal (not only cognitive processes). rational precautions: A review of the psychology of fear appeals. Psychology and Health, 16 (6), 613-630. REFERENCES Shah, JY., Friedman, R., & Kruglanski, AW. (2002). Forgetting all Forster, J., Liberman, N., & Kuschel, S. (2008). The effect of else: On the antecedents and consequences of goal shielding. global versus local processing styles on assimilation versus Journal of Personality and Social Psychology, 83 (6), 1261- contrast in social judgment. Journal of Personality and Social 1280. Psychology, 94(4), 579-599. Smith, K. & Stutts, MA. (2003). Effects of short-term cosmetic Hoyle, R., & Sherrill, M. (2006). Future orientation in the self- versus long-term health fear appeals in anti-smoking system: possible selves, self-regulation, and behavior, Journal advertisements on the smoking behavior of adolescents, of Personality, 74(6), 1673-1696. Journal of Consumer Behavior, 3(2), 157-177. Manyiwa, S. & Brennan, R. (2012). Fear appeals in anti-smoking Wakslak, C., Nussbaum, S., Liberman, N., & Trope, Y. (2008). advertising: How important is self-efficacy?,Journal of Representations of the self in the near and distant future, Marketing Management, 28(11/12), 1419-1437. Journal of Personality and Social Psychology, 95(4), 757-773. Morales, A., Wu, E. & Fitzsimons, G. (2012). How disgust enhances the effectiveness of fear appeals,Journal of Marketing Research, 49(June), 383-393. Passyn, K. & Sujan, M. (2006). Self-accountability emotions and fear appeals: Motivating behavior, Journal of Consumer Research, 32(2), 583-590. Rogers, RW. (1975). A protection motivation theory of fear appeals and attitude change. J.Psych 91: 93-114. How Does Power Distance Belief Affect Consumers’ Preference for User- versus Designer-Designed Products? Xiaobing Song, Dalian University of Technology, China Yinlong Zhang, University of Texas at San Antonio, USA Vikas Mittal, Rice University, USA

EXTENDED ABSTRACT CI: [-1.19, -.02]) and trust (95% CI: [-.50, -.03]) at different levels User design, whereby firms draw on their user communities to of PDB. For low-PDB condition, identification (95% CI: [.03, .92]) generate ideas for new products (Dahl, Fuchs, and Schreier 2015), mediated the positive effect of user design on purchase intention, has become increasingly popular among marketers. However, there whereas trust (95% CI: [-.27, .05]) did not. For high-PDB condition, are substantial differences in the effectiveness of this design approach the negative relationship between user design and purchase intention across countries, suggesting that cultural orientation is important in was mediated by trust (95% CI: [-.55, -.12]) but not by identification understanding the effect of user design on consumer preferences. (95% CI: [-.48, .27]). Thus, hypothesis 2a and hypothesis 2b were Until now, the extant research has not yet investigated the role of supported. cultural orientation. We propose that power distance belief (PDB), Study 4 was a 4 (Design philosophy: novice designer design defined as “the extent to which a society accepts and views as inevi- vs. designer design vs. user design vs. male user design) between- table or functional human inequality in power, wealth, or prestige” subjects design. The description in “designer design” and “user de- (Oyserman 2006, 353), moderates the relationship between design sign” conditions was identical to Study 3. To manipulate the female philosophy and consumer preference. More specifically, we predict participants’ identification with the user-driven company, we addi- that consumers with high (low) PDB will prefer designer- (user-) tionally described the members of the user-community as predomi- designed products to user- (designer-) designed products (H1). Ad- nantly male in the “male user design” condition. We manipulated ditionally, the impact of low PDB on preferring user-designed prod- participants’ trust in the designer-driven company by undermining ucts to designer-designed products is mediated by a stronger iden- the expertise of the internal designers in the “novice designer design” tification with the user-driven company (H2a), while the impact of condition. We measured purchase intention, identification, trust and high PDB on preferring designer-designed products to user-designed chronic PDB. The results revealed a significant interaction effect be- products is mediated by a stronger trust in the designer-driven com- tween design philosophy and chronic PDB (t(368) = -3.26, p = .001). pany (H2b). More importantly, when identification was manipulated to be low, Study 1 used Apache’s country-level market share to measure there was no significant difference in purchase intention between the consumers’ preference between user- and designer-designed prod- low-PDB participants in the “male user design” and the “designer de- ucts. We used Hofstede’s country scores of power distance (PD) as sign” conditions (Mmale user = 4.13 vs. Mdesigner = 4.18; t(368) = -.16, p > independent variable, while individualism, masculinity, uncertainty .80), suggesting that identification was underlying low-PDB’s effect avoidance, long-term orientation per capita GDP, educational index, on preference between designer and user designed products. How- and the Internet users index of each country as control variables. The ever, when trust was manipulated to be low, no significant difference results of regression on 75 matching countries indicated that only PD was found between the high-PDB participants’ purchase intention in had a significant negative effect on Apache’s market share (b = -.009, the “novice designer design” and “user design” conditions (Mnovice z = -2.21, p = .02), such that Apache’s market share was smaller in designer = 4.47 vs. Muser = 4.49; t(368) =- .07, p > .90), suggesting that countries with higher PD than in those with lower PD. Further, we trust was mediating high-PDB’ effect on dependent variable. Thus, found that the relationship between PD and user-designed product this study provided stronger evidence regarding the underlying psy- preference is robust by controlling for alternative explanations such chological processes using a moderation-by -process design. as other cultural dimensions and economic variables. Taken together, This paper contributes to the relevant literature in several sig- the results supported hypothesis 1 at country level. nificant ways: advancing user-design research by uncovering a very Study 2 was a 2 (PDB: low US vs. high China) × 2 (Design important theoretical boundary condition, i.e., PDB; documenting philosophy: designer design vs. user design) between-subjects de- a new psychological process for user-design evaluation by show- sign. The results revealed a significant interaction between country ing that trust is also essential to explain consumer’s preference for and design philosophy (F(1, 463) = 24.67, p < .001). American par- user- versus designer-designed products among high-PDB consumers; advancing the PDB literature by demonstrating that PDB effect ticipants (low PDB) demonstrated a stronger preference (Muser A = .43) for user-designed products than for designer-designed products can be mediated by different psychological process, depending on the decision contexts. (Mdesigner A = -.18; F(1, 463) = 10.31, p = .001), consistent with Dahl et a. (2015). The Chinese participants (high PDB) indicated a reverse REFERENCES pattern (Muser A = -.36, Mdesigner A = .39; F(1, 463) = 14.46, p < .001). Thus, hypothesis 1 was supported at individual level. Dahl, Darren W., Christoph Fuchs, and Martin Schreier (2015), Study 3 was designed to investigate the underlying mechanism “Why and When Consumers Prefer Products of User-Driven proposed in hypotheses 2a and 2b. It was a 2 (PDB: low vs. high) Firms: A Social Identification Account,” Management Science, × 2 (Design philosophy: designer design vs. user design) between- 61 (8), 1978-88. doi:10.1287/mnsc.2014.1999. subjects design. We used the sentence-scrambling task developed by Oyserman, Daphna (2006), “High Power, Low Power, and Equality: Zhang et al. (2010) to prime PDB, and we manipulated design phi- Culture Beyond Individualism and Collectivism,” Journal losophy following Schreier et al. (2012). The results revealed that the of Consumer Psychology, 16 (4), 352–56. doi:10.1207/ interaction effect between design philosophy and PDB (F(1, 172) = s15327663jcp1604_6. 8.45, p = .004) was significant. A planned contrast indicated a similar pattern to Study 2. More importantly, a bootstrap analysis revealed the significant moderated mediation effects of identification (95%

Schreier, Martin, Christoph Fuchs, and Darren W. Dahl (2012), Zhang, Yinlong, Karen Page Winterich, and Vikas Mittal (2010), “The Innovation Effect of User Design: Exploring Consumers’ “Power Distance Belief and Impulsive Buying,” Journal Innovation Perceptions of Firms Selling Products Designed of Marketing Research, 47 (Oct.), 945–54. doi:10.1509/ by Users,” Journal of Marketing, 76 (5), 18-32. doi:10.1509/ jmkr.47.5.945. jm.10.0462. Unethical Product Returning as a Function of Consumers´ Experienced Psychological Distance Anja Spilski, Saarland University, Germany Andrea Groeppel-Klein, Saarland University, Germany Hannah Jungfleisch, Saarland University, Germany Anja Bsdurek, Saarland University, Germany

EXTENDED ABSTRACT Hypothesis 2: Firms size moderates the effect proposed in H1. Unethical product returning refers to a form of consumer mis- More specifically, the effect of an online (vs. of- behavior that is characterized by abuse of a firm’s returns policy by fline) channel on unethical product returning via a consumer. Harris (2008) defined unethical product returning to be social distance is stronger for small firms while knowingly returning “a functional but used or post-purchase dam- it is weaker or even absent for large firms. aged product to any store without informing the store of the use or the true cause of damage” (p. 464). Unethical product returning As a third contribution, we consider ways to decrease unethical by consumers is a source of great expense for retailers in terms of product returning. Previous studies have focused on financial meth- re-shipping, quality checks after the return, repackaging or even ods to reduce unethical product returning, such as fees for (re)ship- destroying the products. Therefore, research and practice warrant ment (Chu, Gerstner, and Hess 1998; Hjort and Lantz 2012). The explanations for the prevalence of this misbehavior (Rosenbaum, present paper supplements the consideration of financial costs by Kuntze, and Wooldridge 2011). taking “psychological cost” (Chu, Gerstner, and Hess 1998, p. 143) Literature previously focused on issues such as which products into account. By suggesting that a smaller psychological distance af- typically are returned (Piron and Young 2000), motivations to de- fects the psychological cost of unethical product returning, we ana- fraud (Schmidt et al. 1999), demographic or psychographic factors lyze possibilities for firms to reduce UPR via reducing psychological that influence unethical product returning (King and Dennis 2003; distance (Study 2). We analyze the impact of “giving the online re- Harris 2008), neutralization techniques of customers who defraud tailer a face” – a personalization strategy that applies a personal pur- (Rosenbaum, Kuntze, and Wooldridge 2011), determinants of the chase advisor on the website. We derive the following hypotheses: success of the fraud (Harris 2010), the perspective of service em- Hypothesis 3: An online retailer can decrease the probability ployees on unethical product returning (O’Brien, Hill, and Autry of unethical product returning by applying (vs. 2009), and contextual factors that influence unethical product return- not applying) a personalization strategy on the ing (e.g., firms’ return policies, Chu, Gerstner, and Hess 1998). website. The effect of personalization is sequen- This article extends previous research on contextual influenc- tially mediated by reduced social distance and es on unethical product returning in terms of the following issues: increased anticipated return risk. First, we introduce and systematically test the impact of psychological distance experienced by consumers towards a company on the Hypothesis 4: The effect proposed in H3 is stronger for small probability of executing unethical product returning. While recent firms and weaker or even absent for large firms. research has shown the importance of psychological distance for the extent of trust associated with a retailer (Darke et al. 2016), we We conducted two experimental studies. Study 1 employed a consider psychological distance in the area of consumer misbehav- 2 (retail channel: offline vs. online store) × 2 (firm size: small vs. ior. Psychological distance is connected to construal-level theory large) between-subjects design. Study 2 focused on the online chan- (Trope, Liberman, and Wakslak 2007). The basic assumption is that nel and used a 2 (personalization strategy: absent vs. present) × 2 more distal entities are constructed on a higher level and are there- (firm size: small vs. large) between-subjects design. We applied a fore represented more abstractly than proximal entities. Abstraction, scenario approach in the context of a jewelry retailer (Study 1) and a then, can influence how consumers process and evaluate information clothing retailer (Study 2). We used third-person scenarios, which are (Dhar and Kim 2007). We focus on social distance as one facet of recommended to decrease socially desirable answering. Participants psychological distance because of its special importance for unethi- indicated the probability of returning the product and experienced cal product returning, as research has found consumers to hesitate to psychological distance from the third-person perspective (Fisher execute unethical product returning when they anticipate embarrass- 1993). Since only participants were included who had recognized the ment in front of service personnel (e.g., Harris 2010). firm’s return policy, a higher probability can be interpreted as abuse Second, we systematically consider two contextual variables of these rules. We used mediation analyses and moderated media- that may influence psychological distance; in particular, we analyze tion analyses to calculate effects. The results confirmed retail chan- how the retail channel (offline vs. online) alone or in interaction with nel to influence consumers’ experienced social distance and conse- firm size (small vs. large firms) affects psychological distance and quently probability of unethical product return (H1). This effect was therefore the probability of unethical product returning (Study 1). We moderated by firm size (H2). We found that personalization strategy derive two hypotheses: on a website reduced social distance. The results confirm a chain of effects of personalization on the probability of unethical product Hypothesis 1: The probability of unethical product returning returning via social distance and anticipated return risk (H3). H4, is higher when the product is purchased online suggesting a moderation of the website personalization effect by firm than when it is purchased at a brick-and-mortar size, was rejected. (offline) store. This relationship is mediated by Several further research avenues are worth considering. First, the social distance experienced by consumers. we focused on unfamiliar retailers. Since research has discussed the relevance of familiarity for social distance (Edwards et al. 2009), further studies should ask whether the effects found here hold for Advances in Consumer Research 896 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 897 familiar retailers. In Study 2, participants were exposed to a purchase Harris, Lloyd C. (2010), “Fraudulent consumer returns: exploiting advisor they did not know. Therefore, research could consider the retailers’ return policies,” European Journal of Marketing, 44 effects of avatars that are used in integrated communication (e.g., (6), 730–747. when consumers already know the avatars from advertising or oth- Hjort, Klas and Björn Lantz (2012), “(R)e‐tail borrowing of party er sources). Second, as the results for H3 have shown, anticipated dresses: an experimental study,” International Journal of return risk served as a mediator. However, processes that are more Retail & Distribution Management, 40 (12), 997–1012. closely related to relational aspects should be considered (as the im- King, Tamira and Charles Dennis (2003), “Interviews of pact of reduced social distance on trust; Darke et al. 2016). deshopping behaviour: an analysis of theory of planned behaviour,” International Journal of Retail & Distribution REFERENCES Management, 31 (3), 153–163. Chu, Wujin, Eitan Gerstner, and James D. Hess (1998), “Managing O’Brien, Matthew, Donna J. Hill, and Chad W. Autry (2009), Dissatisfaction: How to Decrease Customer Opportunism by “Customer Behavioral Legitimacy in Retail Returns Episodes: Partial Refunds,” Journal of Service Research, 1 (2), 140–155. Effects on Retail Salesperson Role Conflict,”Journal of Darke, Peter R., Michael K. Brady, Ray L. Benedicktus, and Marketing Theory and Practice, 17 (3), 251–266. Andrew E. Wilson (2016), “Feeling Close From Afar: The Piron, Francis and Murray Young (2000), “Retail borrowing: Role of Psychological Distance in Offsetting Distrust in insights and implications on returning used merchandise,” Unfamiliar Online Retailers,” Journal of Retailing, 92 (3), International Journal of Retail & Distribution Management, 287–299. 28 (1), 27–36. Dhar, Ravi and Eunice Y. Kim (2007), “Seeing the Forest or the Rosenbaum, Mark S., Ronald Kuntze, and Barbara Ross Trees: Implications of Construal Level Theory for Consumer Wooldridge (2011), “Understanding unethical retail disposition Choice,” Journal of Consumer Psychology, 17 (2), 96–100. practice and restraint from the consumer perspective,” Edwards, Steven M., Jin Kyun Lee, and Carrie La Ferle (2009), Psychology and Marketing, 28 (1), 29–52. “Does Place Matter When Shopping Online? Perceptions Schmidt, Ruth A., Fiona Sturrock, Philippa Ward, and Gaynor of Similarity and Familiarity as Indicators of Psychological Lea‐Greenwood (1999), “Deshopping – the art of illicit Distance,” Journal of Interactive Advertising, 10 (1), 35–50. consumption,” International Journal of Retail & Distribution Fisher, Robert J. (1993), “Social Desirability Bias and the Validity Management, 27 (8), 290–301. of Indirect Questioning,” Journal of Consumer Research, 20 Trope, Yaacov, Nira Liberman, and Cheryl Wakslak (2007), (2), 303. “Construal Levels and Psychological Distance: Effects on Harris, Lloyd C. (2008), “Fraudulent Return Proclivity: An Representation, Prediction, Evaluation, and Behavior,” Empirical Analysis,” Journal of Retailing, 84 (4), 461–476. Journal of Consumer Psychology, 17 (2), 83–95. Why Garlic Ice Cream? Innovative Line Extensions Can Increase Choice of a Brand’s Pre-existing Products Brittney L. Stephenson, Washington University in St. Louis, USA Cynthia E. Cryder, Washington University in St. Louis, USA Stephen M. Nowlis, Washington University in St. Louis, USA Robyn A. LeBoeuf, Washington University in St. Louis, USA

EXTENDED ABSTRACT vors. Replicating our previous findings, choice quantity of the ex- A common marketing strategy, especially recently, is to offer isting items increased when innovative extensions were introduced product line extensions that are innovative and unusual. Crest tooth- (M=2.01) compared to non-innovative extensions (M=1.36, p=.01) paste recently introduced Mint Chocolate, Vanilla Mint and Lime or no extensions (M=1.53, p=.04). Our prediction that innovative ex- Spearmint flavors. A serious concern with this strategy is that line tensions will lead to an increase in perceived brand innovativeness extensions might cannibalize the sales of the original items in the was also confirmed. A test of mediation supports that introducing line. Prior research has found that introducing new extensions often innovative extensions leads to higher perceived brand innovation, leads to cannibalization (Mason & Milne, 1994, Kerin, Harvey & which ultimately increases choice quantity of existing items (95% Rothe, 1978, Copulsky, 1976). CI: [-.30, -.03]). We examine the effect of introducing innovative line extensions Study 3 (n=246) was designed to rule out an alternative expla- on the choice of existing items in the line. Prior work on attribute nation of our results. One could argue that the innovative flavors are transfer has found that perceptions of specific attributes can flow also disgusting, leading to an attraction effect, in which introducing between a brand and its extensions (Keller and Aaker 1992; Park, inferior (or disgusting) options boosts the choice share of similar but Milberg, and Lawson 1991). Furthermore, research on fit would sug- superior options. We test this alternative explanation by altering both gest that line extensions should also transfer perceptions of attributes the relative disgust and relative innovativeness of flavors. between the brand and its line extensions (Aaker and Keller 1990). Participants were randomly assigned to one of four conditions We specifically focus on whether an innovative line extension in a 2 (Innovativeness: low, high) x 2 (Disgust: low, high) between- can transfer the perception of innovativeness back to the parent subjects design. Participants followed the same procedure as study brand, making the brand more appealing. Previous work has found 1A, however, we changed the extension flavors participants saw. Par- that higher perceived innovativeness is often related with more posi- ticipants saw extension flavors either high or low on disgust and high tive brand attitudes (Heath, DelVecchio, and McCarthy 2011). Thus, or low on innovativeness, depending on their condition. We found innovative brands are likely to be more positively evaluated and may a main effect of innovativeness, such that more innovative- exten increase purchase intent. We predict that introducing an innovative sions increased choice quantity of the existing items. Yet there was line extension will lead consumers to purchase a greater number of no main effect of disgust. These results suggest that higher levels of the brands existing items, ultimately reversing cannibalization. innovativeness, and not disgust, increase the choice of existing items Four experiments test our hypothesis. We first show that when in the product line. The interaction was not significant. a brand introduces innovative line extensions (versus non-innovative In study 4 (n=410), we address another alternative explanation, extensions), people choose to purchase more existing items (Studies that if customers find the innovative extensions unappealing for any 1A and 1B). We then show the effect replicates across various prod- other reason, aside from disgust, the extensions may yield a contrast uct categories and is mediated by perceived brand innovativeness effect. To rule out this alternative explanation, we ran a study with a (Study 2). Our last two studies rule out alternative explanations that competitor brand in the choice set to show that only choice share of 1) the effect is driven by disgust (Study 3) and 2) the effect is due to the existing items of the parent brand increases – not choice share of a contrast effect (Study 4). competing brands. The design was identical to study 1A, except that In studies 1A (n=150) and 1B (n=601), participants were ran- a competing brand was offered in the choice set. Scoops Ice Cream domly assigned in a 3-cell (Extension: None, Non-Innovative, Inno- (competitor brand) offered: Vanilla, Chocolate, Strawberry, Butter vative) between-subjects design. All participants saw three existing Pecan, Coffee, Mint. After seeing all items offered by both brands, items offered by a hypothetical brand (Study 1A: Emmi’s) or a na- participants indicated the number of each item (from Emmi’s and tional brand (Study 1B: Breyer’s): Vanilla, Chocolate, and Strawber- Scoops) that they would purchase. ry. In addition to the existing items, participants saw three extension Replicating our main effect, we find that when Emmi’s offers flavors based on their condition. Non-Innovative Extension: Cookies innovative extensions, people choose more existing items (M=1.35) & Cream, Chocolate Chip, and Neapolitan. Innovative Extension: than when no extensions (M=1.08, p=.06) or non-innovative exten- Garlic, Cheddar Cheese, and Avocado Jalapeno. No Extension: no sions are offered (M=0.67,p <.001). People chose more items in total extension items. After reading the list of items, participants indicated (existing + extensions) from Emmi’s when innovative extensions are a purchase quantity for each item. offered (M=2.41) compared to non-innovative extensions (M=1.93, We found that choice quantity of existing items increased sig- p<.01). Most importantly, the total number of Scoops items chosen did not increase when Emmi’s introduced innovative extensions nificantly when an innovative extension is introduced1A (M =2.63, (M=2.62) compared to no extensions (M=2.78, F(1, 407)<1), sug- M1B=2.40) compared to a non-innovative extension (M1A=0.81, gesting that an innovative extension does not benefit all brands (e.g., p<.001); M1B=1.07, p<.001) or no extension (M1A=1.45, p<.001; through a contrast effect), but mainly benefits the parent brand. M1B=1.64, p=.001). Study 2 (n=297) was designed to show our effect replicates in Altogether, we find evidence that innovative line extensions other product categories and to examine if perceived brand innova- may reverse the effect of cannibalization. Innovative extensions in- tiveness mediates this effect. The procedure was identical to study crease perceptions of the brand’s innovativeness, which leads con- 1A, except that we used Kai’s Toothpaste in place of Emmi’s Ice sumers to purchase a higher quantity of the original items in the line. Cream. Toothpaste flavors were substituted for the ice cream fla-

Advances in Consumer Research 898 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 899 REFERENCES Kerin, Roger A., Michael G. Harvey, and James T. Rothe (1978), Aaker, David A. and Kevin Lane Keller (1990), “Consumer “Cannibalism and New Product Development,” Business Evaluations of Brand Extensions,” Journal of Marketing, 54 Horizons, 21 (October), 25-31. (1), 27-41. Mason, Charlotte H. and George R. Milne (1994), “An approach Copulsky, William (1976), “Cannibalism in the Marketplace,” for identifying cannibalization within product line extensions Journal of Marketing, 40 (October), 103-105. and multi-brand strategies,” Journal of Business Research, 31 Heath, Timothy B., Devon DelVecchio and Michael S. McCarthy (2-3), 163-170. (2011), “The Asymmetric Effects of Extending Brands to Park, Whan C., Sandra Milberg, and Robert Lawson (1991), lower and Higher Quality,” Journal of Marketing, 75 (4), 3-20. “Evaluation of Brand Extensions: The Role of Product Feature Keller, Kevin Lane and David A. Aaker (1992), “The Effects of Similarity and Brand Concept Consistency,” Journal of Sequential Introduction of Brand Extensions,” Journal of Consumer Research, 18 (2), 185-193. Marketing Research, 29 (1), 35-50. The Name Game: How Naming Promotes Effects Beneficial to Marketers Jennifer L. Stoner, University of North Dakota, USA Barbara Loken, University of Minnesota, USA Ashley Stadler Blank, University of St. Thomas, USA

EXTENDED ABSTRACT nificantly higher than all the other conditions individually (all p’s < Toyota is encouraging car owners to name their cars and has pro- .10). For the branded condition, none of the contrasts were significant vided close to 100,000 name badges for owners to affix to their autos (all p’s > .10). Once again, the naming effect was significantly medi- (Diaz 2015). Yet, the phenomenon of consumers naming even nonhu- ated by psychological ownership. Potential alternative mediators such man-like products has not been examined by academic researchers. as attachment, self-efficacy, and involvement were not significant. When a person spontaneously names an inanimate object (such as a Study 4 attempted to understand what name features drive self- car), the person is probably doing so because he likes the object. But name success. We employed a mixed design where product (stapler, what about the inverse: what happens when a company asks or directs mug) was a between subjects factor and name condition (no name, the consumer to name an object? In the research reported here, we assigned descriptive name, assigned non-descriptive name) was hypothesize and find that naming of products bestows benefits to the a within subjects factor (p’s > .05). There was no difference in the products when consumers are directed or requested to name them. product conditions and so these data were collapsed for analysis. A We hypothesize that naming of an object may increase liking repeated measures ANOVA on attitudes (F(2, 240) = 58.10, p < .01) for it because it increases feelings of ownership. Naming is an activ- and purchase intentions (F(2, 240) = 55.76, p < .01) revealed a sig- ity generally performed for newborns or pets in one’s family, or for nificant main effect of name. Participants reported higher attitudes nonhuman objects (e.g. dolls, cars, stuffed animals) that the namer (Mself = 5.34) and purchase intentions (Mself = 4.84) for self-names believes to “be mine”. Abundant research on the endowment effect than assigned descriptive names (attitudes: Mdescriptive = 4.26; purchase has shown that when consumers take ownership of an object, it in- intentions: Mdescriptive = 3.89) or non-descriptive (attitudes: Mnon-descriptive creases in value (e.g. Kahneman, Knetsch, and Thaler 1990). These = 3.36; purchase intentions: Mnon-descriptive = 2.98) names. Participants endowment or ownership effects are also found when consumers feel also reported higher attitudes and purchase intentions for descriptive psychological (rather than actual) ownership. (Pierce, Kostova, and versus non-descriptive names (all p’s < .05). Furthermore, the ME- Dirks 2003). MORE macro (Montoya and Hayes 2017) for repeated measures de- Study 1 explores the basic effect of naming and whether it per- signs demonstrated that this effect was driven by psychological own- sists over time. Student participants were recruited for a two-part ership which in turn was driven by the fit and creativity of the name. study on stress reduction. Participants were randomly assigned to Study 5 compares the effects of naming to a form of co-creation, either a naming or no-name condition. During the first lab session, designing a mug. Student participants were randomly assigned to one all participants received a plain yellow stress ball and were either in- of three conditions: naming, no-name, and designing. In the design structed to name it or not. Four weeks later, participants were asked to condition, participants were given colored pencils to design a mug. report how much someone would have to pay them to buy their stress All participants completed the same scales used in Study 3. An ANO- ball from them. As predicted, the naming condition placed greater VA on attitudes revealed a significant main effect (F(2, 101) = 10.37, value on their stress ball than those in the no-name condition, F(1,38) p < .01). Participants reported higher attitudes for self-design (M = = 4.42, p < .05. In fact, price values were remarkably more than $1 5.23) than self-name (M = 4.48) or the control (M = 3.61). Impor- higher for participants in the naming than no-name condition (M = tantly, participants still reported higher attitudes for self-name versus $4.07 vs. $2.67). the control (all p’s < .05). Once again, this effect was mediated by In Study 2 participants were randomly assigned to one of four psychological ownership. Interestingly, affective commitment also conditions: naming, no-name, assigned descriptive name “Blue”, or mediated the positive impact of self-design but did not mediate the assigned non-descriptive name “Steve”. All participants viewed a pic- difference between self-name and the control condition. ture of a stapler. Participants in the naming condition were asked to provide a name for it whereas in the assigned name conditions the REFERENCES stapler had already been given a name. An ANOVA revealed a signifi- Diaz, Ann-Christine (2015), “Toyota’s ‘Mas que un Auto’ Is More cant effect of condition, F(3, 196) = 3.75, p = .01, on purchase inten- than Your Average Campaign: Hispanic ‘Badge’ Campaign via tion. Purchase intentions were highest in the naming condition (M = Conill Gives Real Individuality to Cars,” Advertising Age, < 4.49), followed by the “Blue” (M = 3.75), “Steve” (M = 3.74) and no http://adage.com/article/behind-the-work/toyota-s-mas-auto- name (M = 3.50) conditions. Planned contrasts showed that purchase average-campaign/297904/> last accessed August, 22, 2016. intent was higher in the naming condition than each individual condi- Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler (1990), tion (p’s < .05). Additionally, mediation analysis demonstrated that “Experimental Tests of the Endowment Effect and the Coase psychological ownership mediated this effect. Theorem, Journal of Political Economy, 1325-48. In Study 3 we compare the effects of naming for an unbranded Montoya, Amanda K., and Andrew F. Hayes (2017), “Two- product (laptop) and for a brand product with a strong prior brand Condition Within-Participant Statistical Mediation Analysis: image (Apple laptop). Participants were randomly assigned to a 2 A Path-Analytic Framework.,” Psychological Methods, 22 (1), (brand, no brand) X 4 (naming, no-name, assigned descriptive name 6-27. “Slimbook”, or assigned non-descriptive name “George”) between- Pierce, Jon L., Tatiana Kostova, and Kurt T. Dirks (2003), “The subjects design. Participants reported their attitudes toward the laptop State of Psychological Ownership: Integrating and extending on five items. Results revealed in the unbranded condition favorabil- a Century of Research,” Review of General Psychology, 7(1), ity was highest in the naming condition (M = 5.03), followed by the 84-107. “Slimbook” (M = 4.63), no name (M = 4.51) and “George” (M = 4.44) conditions. Planned contrasts showed the naming condition was sig-

Advances in Consumer Research 900 Volume 45, ©2017 The Influence of Visual Aesthetics on Food Choice: The Moderating Effect of Food Processing Jacob Suher, Portland State University, USA Courtney Szocs, Portland State University, USA Koert Van Ittersum, University of Groningen, The Netherlands

EXTENDED ABSTRACT a variable texture and color due to the presence of apple chunks and Each year billions of pounds of produce are thrown away due skin. The lateral position of the options was counterbalanced. After to aesthetic imperfections (i.e., blemishes, unconventional shapes; viewing the pair of foods participants indicated their preference (1 Bratskeir 2015), often because retailers and consumers refuse to = definitely option A, 7 = definitely option B). The data was pre- purchase products that fail to meet strict cosmetic standards for vi- pared so that “option A” always represented the aesthetically imper- sual perfection (Royte 2016). Interestingly, as safe-to-eat fruits and fect food option. A one-way ANOVA revealed that participants had vegetables are discarded because of their imperfect appearance, a stronger preference for the aesthetically perfect option when they food manufacturers and restaurant chains intentionally manufacture viewed unprocessed (vs. processed) foods (Mprocessed = 4.3056 vs. Mun- products that are aesthetically imperfect. For example, employees at processed = 6.1714; F (1, 69) = 25.587, p < .01). Domino’s Pizza purposefully shape dough into irregular rectangles Consumers rely on heuristics when information is ambiguous when preparing Artisan Pizza (Choi 2013). Why is it that retailers (Kruger et al. 2004), but are less likely to use such heuristics when and manufacturers discard fruits and vegetables that contain aes- information is unambiguous. Thus, in the context food choices, the thetic imperfections yet design processed foods to contain aesthetic effect of aesthetic imperfections should be less influential in driving imperfections? choice when taste information is available (i.e., unambiguous). Study In contrast to conventional wisdom that “beautiful is good” 2a tested this idea using a one factor between subjects experiment (Veryzer and Hutchinson 1998), our research provides evidence that (taste information present vs. absent). Participants (N=71) chose be- in some cases consumers prefer aesthetically imperfect (vs. perfect) tween an aesthetically perfect and imperfect unprocessed food (i.e., foods. We show that preference for aesthetically perfect (vs. imper- whole oranges). The key finding was that participants were signifi- fect) food depends on the level of mechanical processing the food cantly more likely to choose the aesthetically imperfect orange when has undergone. Specifically, in four studies we show that when given they had the opportunity to sample prior to choosing than when they a choice between unprocessed foods (e.g., two apples) consumers did not have the opportunity to sample prior to choosing (Psampling pres- 2 prefer the aesthetically perfect option. However, when given the ent = 53.66% vs. Psampling absent = 16.67%; χ = 10.059, p < .01). Study choice between two processed foods (e.g., two bowls of applesauce) 2b was similar to study 2a except that the study was a field experi- consumers prefer the aesthetically imperfect option. These effects are ment at a grocery store and shoppers (N = 41) choose between two attenuated for both unprocessed and processed foods when individu- processed foods (i.e., cookies) when taste information was present als have the opportunity to sample the items before choosing. or absent. The results showed that preference for the aesthetically First, we conducted a pretest to identify images of unprocessed imperfect cookie decreased when taste information was present (i.e., and processed foods that differed in terms of the degree of visual when individuals were allowed to sample the food before choice) 2 perfection, but were similar in terms of healthiness, size, and safety. (Pmulti-sensory absent = 71.7% Pmulti-sensory present = 54.0%; χ = 3.461, p = .07). In the main experiment (study 1a), members of an online panel were Collectively, studies 2a and 2b suggest that retailers can overcome randomly assigned to view either a pair of unprocessed foods (i.e., the effects of aesthetic imperfections on food choice by providing carrots) or a pair of processed foods (i.e., pizzas). For each pair, one diagnostic taste information. option was aesthetically perfect and the other option was aesthetical- The results of four studies show that consumers prefer aestheti- ly imperfect (as determined by the pretest). Participants were asked cally perfect (vs. imperfect) unprocessed foods and aesthetically im- to indicate which item they would prefer to eat. The results showed perfect (vs. perfect) processed foods. These effects are attenuated that participants who viewed the pair of unprocessed foods were sig- for both unprocessed and processed foods when individuals have nificantly more likely to choose the aesthetically perfect option than the opportunity to sample the items before choosing. Theoretically, our findings contribute novel evidence that the “form” of a product participants who viewed the pair of processed foods (Pcarrots = 79.59% 2 moderates the influence of visual aesthetics on choice. While much vs. Ppizza = 58.82%; χ = 5.039, p < .05). One shortcoming of study 1a is that we used two different food is known about the stimulus factors and individual differences that items for the processed and unprocessed foods. In study 1b we ad- drive aesthetic impressions (e.g., Veryzer & Hutchinson 1998), we dress this shortcoming by using the same food item and varying only demonstrate that visual aesthetics can have either a positive or, coun- the level of mechanical processing. Study 1b had a one-factor be- terintuitively, a negative influence on consumer choice. Our findings tween subjects design with two experimental conditions (food type: also contribute evidence that food processing influences consumer processed vs. unprocessed). Members of an online panel (N = 71, food choices (Szocs and Lefebvre 2016). From a sustainability perspective, our findings provide evidence that multi-sensory informa- Mage = 30.30; 52.1% females) were randomly assigned to view a pair of apples (unprocessed condition) that were the same weight tion can increase consumers’ acceptance of visually imperfect pro- and type but varied in appearance (i.e., one had a uniform shape and duce. texture and the other had a variable shape and texture). Participants in the processed condition viewed two bowls of applesauce. The REFERENCES applesauce was created by mechanically processing (i.e., blending) Bratskeir, Kate (2015), “Six Billion Pounds of Perfectly Edible an apple and water. Each bowl contained the same amount of apple- Produce is Wasted Each Year, Simply Because It’s Ugly,” sauce, but the bowls varied in appearance. The aesthetically perfect http://www.huffingtonpost.com/2015/05/19/food-waste-ugly- applesauce had a uniform texture and color due to a lack of apple fruits-and-vegs-dont-judge_n_7309432.html chunks and apple skin. The aesthetically imperfect applesauce had

Choi, Candice (2013), “Food Companies Work to Make it Look Szocs, Courtney and Sarah Lefebvre (2016), “The Blender Effect: Natural,” http://www.sandiegouniontribune.com/sdut-food- Physical State of Food Influences Healthiness Perceptions and companies-work-to-make-it-look-natural-2013jun17-story. Consumption Decision,” Food Quality and Preference, 54, html 152-159. Kruger, Justin, Derrick Wirtz, Leaf Van Boven, T.William Altermatt Veryzer, Robert W. and J. Wesley Hutchinson (1998), “The (2004), “The Effort Heuristic,”Journal of Experimental Social Influence of Unity and Prototypicality on Aesthetic Responses Psychology, 40(1), 91-98. to New Product Designs,” Journal of Consumer Research, Royte, Elizabeth (2016), “How ‘Ugly’ Fruits and Vegetables Can 24(4), 374-385. Help Solve World Hunger,” National Geographic, March 2016. Punishing Politeness: Moderating Role of Belief in Just World on Severity? Aparna Sundar, University of Oregon, USA Wendy Paik, University of Oregon, USA

EXTENDED ABSTRACT vate cynicism can be described by the belief that firms are disguising Three studies demonstrate that politeness in advertising affect their own self-interest with altruism (Helm 2004; Odou and de Pech- how consumers punish brands for unethical behavior. Those low in peyrou 2011). Cynical consumers seem to be motivated by a desire belief in just world (BJW) were more likely to punish a brand when to punish companies whom they distrust (Helm 2004). Trust is the more polite language was used. Both cynicism and trust mediated the belief that the motives of others will be beneficial to one’s interests effect of ethical practice on level of punishment. (Turner and Valentine 2001). Brand trust is the ability of a brand to elicit trust from consumers (Chaudhuri and Holbrook 2001). Trust- Introduction ing consumers seem to be motivated by a desire to reward trustwor- People are more likely to apply moral judgements to psycho- thy companies (Helm 2004). logically distant situations (Eyal et al. 2008). Psychological distance can be influenced through language. One aspect of language, polite- Study 1 ness, has been shown to increase social distance between speakers The objective of Study 1 was to evaluate whether consumers (Brown and Levinson 1987; Holtgraves and Yang 1990; 1992). The would penalize brands differently based on the politeness demon- purpose of this research is to understand the effect of politeness on strated in an advertisement by the brand. moral judgements of the actions of brands. Design and Participants We show that polite language in advertising can affect consum- Participants (N = 124; 54.8% female; M = 35.79) were ran- ers’ intentions to punish a brand whose actions are ethically ambigu- age domly assigned to one of three conditions (politeness: control vs. ous. We explore the role of belief in a just world (BJW) when making less polite vs. more polite). Participants were asked to review an such judgments. In addition, we also explore the underlying mecha- advertisement for a fictitious apparel brand called CJ. In the more nism of trust and cynicism that drives punishment and moral percep- polite condition, the tag line was the same as in the control condition tion of the brand. followed by the word sir, and in the less polite condition, was followed by the word dude. Participants were then randomly presented Conceptual Development with one of two press reports. They were that CJ was recently in the Politeness news and that a press report stated either: CJ purchases clothes from Formal titles, e.g. sir or Mister, are considered more polite than manufacturers that do not use child labor, however the manufactures less formal terms of address, e.g. dude or mate (Ambady et al. 1996; do not provide health care benefits to their employees (pretested M Brown and Levinson 1987; Clark and Schunk 1980; Holtgraves and = 3.93, SD = 1.34) or that CJ uses low impact dyes, but the colors Yang 1990; 1992). Polite speech, however, increases the percep- fade in clothing and are not durable (pretested M = 4.25; SD = 1.54). tion of social distance between speakers (Brown and Gilman 1989; These two statements were not significantly different on ethical rat- Brown and Gilman 1960; Holtgraves and Yang 1990; 1992). Situa- ing (p = .38). tional factors can also cause otherwise polite speech to be considered Procedures and Measures manipulative, sarcastic, or even offensive (Watts 2003). Participants were asked to indicate how much they agreed The greater the psychological distance, the more abstract peo- with the statement, “I feel that CJ did something morally wrong,” ple’s thinking (Trope and Liberman 2010). More polite speech is as- (anchored: 1 = strongly disagree; 7 = strongly agree; adapted from sociated with a higher level of psychological distance (Stephan et al. Cheng, Ottati, and Price, 2013). Then they were asked to, “Rate the 2010). Judgements of morality are also associated with psychologi- degree of punishment for such behavior,” (anchored: 1= lenient; 7 cal distance. The greater the social distance between parties involved = severe; adapted from Cheng, Ottati, and Price 2013). Participants in a moral transgression, the more morally wrong an action is per- then completed the sixteen-item measure of BJW scale (and Peplau ceived (Eyal et al. 2008). It therefore follows that more polite speech, 1973). which increases social distance, should also increase the perception that a moral transgression is wrong. Results An ANCOVA was conducted with politeness in the advertise- Belief in a Just World ment and the composite score of BJW (α = .90) as predictors of pun- BJW is a belief that people get what they deserve (Lerner 1980). ishment. There was a significant effect of politeness in advertising on When encountering an injustice, BJW motivates us to punish a vic- severity of punishment (F(2, 118) = 3.42, p < .05, η2 = .055). There timizer (Lerner 1980). Those high in BJW are more likely to pun- was also a significant effect of BJW on severity of punishment (F(1, ish those who violate social norms (Zhu et al. 2012). Those low in 118) = 6.05, p < .05, η2 = .049). There was a significant interac- BJW perceive others as more motivated by self-interest, and that tion between politeness in the advertisement and the BJW (F(2, 118) they believe that people often exploit others to their advantage (Lip- = 3.09, p < .05, η2 = .05). A spotlight analysis revealed that when kus 1992). BJW has also been shown to be related to psychological BJW was high, the punishment judgments did not vary significantly distance, such that as psychological distance increases, so too does among the three advertisement conditions (p = .67). The control (M BJW (Warner et al. 2012). As such, we believe that when more po- = 3.60, SD = 1.34) did not vary from the high politeness condition lite speech is used, social distance increases, so too does the need to (M = 3.00, SD = 2.55, p = .65), which in turn did not vary from the restore justice. low politeness condition (M = 4.00, SD = 1.73). It was only when the Cynicism and Trust BJW was low that participants reduced the severity of punishment Cynicism and trust are two important dimensions of moral judg- for the control (M= 1.40, SD = .54) and in comparison to the more ment (Turner and Valentine 2001). The goals or values which moti- polite condition (M = 3.63, SD = 2.20; F(1, 118) = 4.80, p < .05). The

Advances in Consumer Research 903 Volume 45, ©2017 904 / Punishing Politeness: Moderating Role of Belief in Just World on Severity? punishment was low even for the less polite condition as well (M = seems plausible that when individuals believe that people get what 1.67, SD = 1.21) in comparison to the more polite condition (M = they deserve, the effectiveness of punishment would be diminished. 3.63, SD = 2.20; F(1, 118) = 3.83, p < .05). This result was despite By the same token those low in BJW, would probably view the ad- the fact that the interaction of BJW and politeness in the advertise- vertising strategy of adopting politeness as an attempt at misleading ment on morality judgments was not significant p( = .22). consumers.

Discussion Study 3 Those low in BJW indicated higher punishment for the brand The objective of this study was to evaluate the role of both which used more polite language over the brand with less polite lan- brand trust in a brand and the consumer cynicism in the observed guage and the control. Thus, our hypothesis that the consumers will condemnation by consumers. To evaluate this, we adopted ethical penalize brands differently based on the politeness of the language statements that were judged to be either low or high on ethical rat- used in an advertisement was supported. ings in the pretest. Study 2 Design and Participants

The objective of Study 2 was to evaluate whether the penalty Participants (N = 162; 56.2% female; Mage = 37.7) were ran- imposed on different brands would manifest in the fine amounts. domly assigned to one of two conditions (ethical rating: high vs. low). A known brand (GEICO) was selected for this study. Partici- Design and Participants pants were told that GEICO was recently in the press. In the low Participants (N = 167; 56.9% female; M = 34.08) were ran- age ethical condition, participants read: A press report noted that GEICO domly assigned to one of two conditions (levels of politeness in ad- hires women who work in cooperatives, however, these women are vertising: more polite vs. less polite). Participants were asked to re- sometimes criminals who don’t pay their taxes (pretested M = 2.77; view an advertisement of the fictitious coffee brand called Jaunt. The SD = 1.17). Participants in the high ethical condition read: A press tag line used in the advertisement varied on two levels of politeness. report noted that GEICO helps disadvantaged communities, but the In the more polite condition, participants read Are you going to have help is only through volunteer service and not monetary in nature a cup of coffee? In the less polite condition, participants read Gonna (pretested M = 5.16; SD = 1.41). These two statements were signifi- have a cup of coffee? Participants were then told: Recently, Jaunt cantly different on ethical ratingt (58.01) = 7.22, p < .001. was in the press. The press report noted that: Jaunt collaborates with local farmers in Peru, but they do not contribute to infrastructure Procedures and Measures improvement in Peru (pretested M = 3.64; SD = 1.51). Participants After reviewing the advertisement and the press statement, were then asked to quantify the punishment levied against the brand. participants were asked to indicate the degree to which GEICO did something morally wrong on the same scale as used in Study 1. Par- Procedures and Measures ticipants then rated the degree of punishment of GEICO. Participants Participants were asked If the Peru government had to levy a were also asked to provide ratings to capture cynicism on a two-item fine on Jaunt what would it be? Participants were asked to indicate scale: I feel cynical about GEICO; I feel skeptical about GEICO’s the fine for the company on a $0 to $250,000.00 sliding scale (modi- business practices (anchored: 1 = strongly disagree; 7 = strongly fied from previous research where consumers levied fines for com- agree). Then brand trust was captured using three items: I trust GEI- pany misconduct [e.g. Ashton-James and Tracy 2012; Rosenblatt et CO; GEICO is an honest brand; GEICO is a safe brand (anchored: al. 1989]). Following this, participants were asked the BJW scale as 1 = strongly disagree; 7 = strongly agree; adapted from Chaudhuri used in Study 1. and Holbrook 2001). Results Results Hypothesis Testing. An ANCOVA was conducted with polite- Morality and Punishment. An analysis of the effect of ethical ness in the advertisement and the composite score of BJW (α = .89) practice on perception that GEICO did something morally wrong as predictors of fines levied by participants. There was a significant was higher when presented with a less ethical practice (M = 4.06, SD effect of politeness in advertising the fine levied (F(1, 163) = 5.34, = 1.53) in comparison to the more ethical practice (M = 2.54, SD = p < .05, η2 = .032). There was also a significant effect of BJW on 1.50; t(159.94) = 6.34; p < .001). An analysis of the effect of ethical severity of punishment (F(1, 163) = 3.75, p < .05, η2 = .023). There practice on severity of punishment yielded a significant difference was a significant interaction between politeness in the advertisement with participants wanting more punishment for the less ethical prac- and BJW (F(1, 164) = 3.92, p < .05, η2 = .025). A spotlight analysis tice (M = 3.94, SD = 1.33) in comparison to the more ethical practice revealed that when BJW was high, the fine did not vary significantly (M = 2.07, SD = 1.50; t(160) = 8.06; p < .001). among the two levels of politeness (p = .96). Specifically, the fine Mediation Analysis. To determine whether cynicism or trust levied in the more polite condition (M = $10,626.63, SD = 29,806.24) accounted for participants’ perception of GEICO, we conducted a did not vary from the less polite condition (M = $10,138.91, SD = mediation analysis (Hayes 2012, Model 4; bootstrapped with 5,000 28869.13). It was only when BJW was low that the fine levied was draws). The analysis revealed that although the ethical practice influ- low for the less polite condition (M = $23,614.33, SD = 38,476) in enced both cynicism and trust, only cynicism mediated the relation- comparison to the more polite condition (M = $240,904.92, SD = ship between how ethical GEICO’s practice was and participant’s 433,574.67; F(1, 163) = 7.75, p < .01). This result was even though perception of how moral GEICO was (95% CI = [–.42, –.04]) but the interaction of BJW and politeness in the advertisement on moral- trust did not mediate the relationship between how ethical GEICO’s ity judgments was not significant p( = .34). practice was and participant’s perception of how moral GEICO was Discussion (95% CI = [–.02, .22]). On the other hand, when the same mediation Those high in BJW showed no significant difference in the fine analysis (Hayes 2012, Model 4; bootstrapped with 5,000 draws) was as a measure of punishment between two advertisements with varied conducted, both cynicism (95% CI = [–.50, –,06]) and trust (95% politeness levels in the language. Those low in BJW, however, indi- CI = [.01, .31]) mediated the effect of ethical practice on level of cated a higher fine for the brand which used more polite language. It punishment. Advances in Consumer Research (Volume 45) / 905

Discussion. The participants of Study 3 indicated that GEICO Hayes, Andrew F. (2012), “PROCESS: A Versatile Computational did more morally wrong when the brand was presented with the less Tool for Observed Variable Mediation, Moderation, and ethical practice than with the more ethical practice. They punished Conditional Process Modeling,” white paper, available at GEICO more for their less ethical practice compared to the more http://www.afhayes.com/public/process2012.pdf . ethical practice. The result indicated that the participants felt more Helm, Amanda (2004),”Cynics and Skeptics: Consumer cynical when they were exposed to GEICO performing a less ethi- Dispositional Trust,” in NA – Advances in Consumer Research cal practice in comparison to the more ethical practice. Individuals Volume 31, eds. Barbara E. Kahn and Mary Frances Luce, trusted GEICO more when presented with a more ethical practice Valdosta, GA: Association for Consumer Research, 345–351. than a less ethical practice. Mediation analyses showed that cyni- Holtgraves, Thomas, and Yang Joong-nam (1990), “Politeness as cism was the only mediator of the relationship between the ethical Universal: Cross-cultural Perceptions of Request Strategies practice and moral perception. Both trust and cynicism were found to and Inferences Based on Their Use,” Journal of Personality be significant mediators of the relationship between ethical practice and Social Psychology, 59 (4), 719–729. of the brand and punishment. ______(1992), “Interpersonal Underpinnings of Request Strategies: General Principles and Differences Due to Culture General Discussion and Gender,” Journal of Personality and Social Psychology, We add on work by Eyal et al. (2008) by showing that polite- 62 (2), 246–256. ness increases social distance and increased social distance leads to Langreth, Robert, and Matthew, Herper (2010), “The Planet Versus greater judgements of wrongdoing. Based on past research on BJW, Monsanto,” Forbes Magazine, December 29. our findings show that the language used in a brand’s advertising Lerner, Melvin J. (1980), “The Belief in a Just World,” In The can potentially harm the brand when it is undergoing a brand crisis Belief in a Just World, Springer US, 9–30. (i.e., an ambiguously unethical business practice reported by the me- Lipkus, Isaac M. (1992), “A Heuristic Model to Explain dia). Individuals in our studies consistently showed that they would Perceptions of Unjust Events,” Social Justice Research, 5 (4), more likely punish a brand which used more polite language in their 359–384. advertising compared to the brand which used less polite language. Odou, Philippe, and Pauline de Pechpeyrou (2011), “Consumer Our results showed that the influence of the politeness of the lan- Cynicism: From Resistance to Anti-consumption in a guage used in advertising was only significant for the individuals Disenchanted World?,” European Journal of Marketing, 45 with low BJW. These results correspond to past findings that those (11/12), 1799–1808. low in BJW believe people often exploit others to their advantage, Rosenblatt, Abram, Jeff Greenberg, Sheldon Solomon, Tom particularly those of higher power and status (Lipkus 1992). Thus, Pyszczynski, and Deborah Lyon (1989), “Evidence for Terror a possible explanation would be that those low in BJW were more Management Theory: I. The Effects of Mortality Salience cautious of the actions of the brand and thus reacted to its unethical on Reactions to Those Who Violate or Uphold Cultural business practice differently. Values,” Journal of Personality and Social Psychology, 57 (4), 681–690. REFERENCES Rublin, Z., and A. Peplau (1973), “Belief in a Just world and Ambady, Nalini, Jasook Koo, Fiona Lee, and Robert Rosenthal Reactions to Another’s Lot: A Study of Participants in the (1996), “More than Words: Linguistic and Nonlinguistic National Draft Lottery,” Journal of Social Issues, 29 (4), Politeness in Two Cultures,” Journal of Personality and Social 73–93. Psychology, 70 (5), 996–1011. Stephan, Elena, Nira Liberman, and Yaacov Trope (2010), Ashton-James, Claire E., and Jessica L. Tracy (2012), “Pride and “Politeness and Psychological Distance: A Construal Level Prejudice How Feelings about the Self Influence Judgments of Perspective,” Journal of Personality and Social Psychology, Others,” Personality and Social Psychology Bulletin, 38 (4), 98 (2), 268–280. 466–476. Sundar, Aparna, and James J. Kellaris (2016), “How Logo Colors Brown, R., and A. Gilman (1960), “The Pronouns of Power and Influence Shoppers’ Judgments of Retailer Ethicality: The Solidarity,” in Style in Language, ed. by Thomas A. Sebeok, Mediating Role of Perceived Eco-friendliness,” Journal of Cambridge: MIT Press, 253–276. Business Ethics, DOI 10.1007/s10551-015-2918-4. ______(1989), “Politeness Theory and Shakespeare’s Four Trope, Yaacov, and Nira Liberman (2010), “Construal-level Theory Major Tragedies,” Language in Society, 18 (02), 159–212. of Psychological Distance,” Psychological Review, 117 (2), Brown, Penelope, and Stephen C. Levinson (1987), Politeness: 440–463. Some Universals in Language Usage, Vol. 4, Cambridge Turner, James H., and Sean R. Valentine (2001), “Cynicism as a University Press. Fundamental Dimension of Moral Decision-making: A Scale Chaudhuri, Arjun, and Morris B. Holbrook (2001), “The Chain Development,” Journal of Business Ethics, 34 (2), 123–136. of Effects from Brand Trust and Brand Affect to Brand Warner, Ruth H., Molly J. VanDeursen, and Anna RD Pope Performance: The Role of Brand Loyalty,” Journal of (2012), “Temporal Distance as a Determinant of Just World Marketing, 65 (2), 81–93. Strategy,” European Journal of Social Psychology, 42 (3), Cheng, Justin S., Victor C. Ottati, and Erika D. Price (2013), 276–284. “The Arousal Model of Moral Condemnation,” Journal of Watts, Richard J. (2003), Politeness, Cambridge University Press. Experimental Social Psychology, 49(6), 1012–1018. Zhu, Luke, Jason P. Martens, and Karl Aquino (2012), “Third Party Clark, Herbert H., and Dale H. Schunk (1980), “Polite Responses to Responses to Justice Failure: An Identity-based Meaning Polite Requests,” Cognition, 8 (2), 111–143. Maintenance Model,” Organizational Psychology Review, 2 Eyal, Tal, Nira Liberman, and Yaacov Trope (2008), “Judging Near (2), 129–151. and Distant Virtue and Vice,” Journal of Experimental Social Psychology, 44 (4), 1204–1209. The Influence of Posture on Taste Evaluations Courtney Szocs, Portland State University, USA Dipayan Biswas, University of South Florida, USA Annika Abell, University of South Florida, USA

EXTENDED ABSTRACT fect of posture on heart rate with heart rate being higher for standing

In some contexts, individuals eat while maintaining a seated (vs. sitting) postures (Msitting = 72.13 vs. Mstanding = 83.28; (F(1, 76) = posture (e.g., at restaurants, in-home dining). However, in other con- 13.13, p < .01). texts individuals eat while maintaining a standing posture (e.g., at If standing is influencing taste by increasing stress then this ef- cocktail parties, grocery store sampling stations). Would eating the fect should persist in the absence of a stress prime, but should be same food while sitting (vs. standing) influence taste evaluations? attenuated when stress is primed due to decreases in perceived taste That is, how does posture influence taste? when individuals are stressed and seated. Study 3 tested this predic- Given that restaurants are beginning to experiment with the tion with a 2 (posture: sitting vs. standing) x 2 (stress: present vs. ab- standing only concept that has long been popular outside of the Unit- sent) between subjects design. One hundred and twenty undergradu- ed States (Fujita 2013), this finding has important practical implica- ate students participated in this study (Mage = 22.62; 40.9% females); tions. In addition, while prior research has examined how the effects however, four individuals did not complete the key dependent mea- of body orientation on consumer decision making much of this work sures leaving a final sample of one hundred sixteen. To manipulate has been done through the lens of embodied cognition (Labroo and stress individuals completed a timed word jumble task (Zellner et al. Nielsen 2010; Strack, Martin and Stepper 1988). We know of no re- 2006). Subsequently, individuals sampled a cookie sitting or stand- search that has examined the effects of posture on taste. ing. The results a 2 (posture) x 2 (stress) ANCOVA with hunger as We build on literature that shows that standing leads to greater the covariate revealed a significant interaction (F(1, 111) = 12.156, physiological stress than sitting (Abalan et al. 1992) and link it with p = .001). In the absence of a stress prime, participants who ate the work that shows that stress decreases the reward value (i.e., liking) cookie when sitting perceived it as better tasting (Msitting = 5.79 vs. of hedonic stimuli (Pizzagalli et al. 2007). We predict that because Mstanding= 4.60; F(1, 112) = 12.63, p < .01). However, when stress was standing increases physiological stress and stress decreases reward primed there was no difference in taste based on whether participants value of hedonic stimuli, consumers will rate the same food as bet- sampled while sitting (vs. standing) (Msitting = 4.74 vs. Mstanding = 4.86; ter tasting if they sample while maintaining a sitting (vs. standing) F(1, 112) = .112, p = .738). posture. We test this hypothesis and the mediating effect of stress in Finally, in Study 4 we wanted to show that relaxation can re- a series of four experimental studies. lieve the stress associated with standing and attenuate the effects of Study 1 examined the basic effect of posture on taste using a posture on taste by increasing taste evaluations when individuals single factor between subjects design (posture: sitting vs. standing). are standing. Study 4 had a 2 (posture: sitting vs. standing) x 2(relaxation: present vs. absent) between subjects design. Posture was Participants (Mage = 24.39; 50% females) sampled a cookie and rated the taste (Elder and Krishna 2010). Participants also rated their manipulated through the presence or absence of chairs. Relaxation physiological stress/tension and hunger levels. We measured hunger was manipulated through the presence of different types of ambient because hunger influences food evaluations (Lozano, Crites and Aik- music (Pham, Hung and Gorn 2011). Participants (N = 111; Mage = man 1999). The results of an ANCOVA with hunger as a covariate 21.86; 64.9% females) sampled and evaluated a cream-filled wafer showed that individuals who ate the cookie while seated perceived it cookie. The results showed that, in the absence of relaxation partici- as better tasting than individuals who ate the cookie while standing pants perceived the cookie as better tasting when they were sitting (M = 5.70 vs. M = 4.79; F(1, 107) = 6.22, p < .05). How- (Msitting = 4.96 vs. Mstanding = 4.22; F(1, 81) = 5.44, p < .05). Tests for sitting standing mediation using Preacher and Hayes’ (2008) 5,000 bootstrap sam- ever, when relaxation was primed taste evaluations when standing ples showed that the indirect effects of posture on taste evaluations increased and there was no significant difference in taste based on through the mediator perceived stress/tension (with hunger included posture (Msitting = 5.48 vs. Mstanding = 5.67; F(1, 107) = .266. p = .607). as a covariate) yielded a confidence interval that did not include zero Collectively, the results of four studies show that posture sys- (B = .1497, SE = .099, 90% CI: .008, .4193) suggesting that stress/ tematically effects taste evaluations by increasing physiological tension mediates the effect of posture on taste. stress. Then, Study 2 replicated the basic effects of posture on taste and also provided physiological evidence that standing postures were REFERENCES associated with greater stress by measuring participants’ heart rate. Abalan F., A. Hui Bon Hoa, D. Guillonneau, E. Ruedas, M. Study 2 had a one-factor between subjects design and was conducted Calache, W. Ellison, F. Rigal, F. Perey, M. Sousselier, and M. in two phases because prior research shows that consuming indulgent Bourgeois (1992), “Effect of Posture on Total Cortisol Plasma foods increase heartrate (Brown et al. 2008). The first phase involved Concentration,” Hormone and Metabolic Research, 24, 595- individuals sampling a cookie while standing (vs. sitting) and then 96. rating the taste. The second phase, which occurred approximately Brown, Clive M., Abdul G. Dulloo, Gayathri Yepuri, and Jean- 3 weeks later involved capturing individuals pulse while they were Pierre Montani (2008), “Fructose Ingestion Acutely Elevates maintaining the same posture they had in the first phase of the study. Blood Pressure in Healthy Young Humans,” American Journal Eighty undergraduate students participated in this study however of Physiology, 294 (3), R730-37. two students did not complete both phases leaving a final sample Elder, Ryan and Aradhna Krishna (2010), “The Effects of Advertising Copy on Sensory Thoughts and Perceived Taste,” of seventy-eight (Mage = 24.01, 47.4% females). An ANCOVA with hunger as a covariate showed that participants perceived the cookie Journal of Consumer Research, 36 (5), 748-56. as better tasting when they ate it while sitting (Msitting = 4.35 vs. Mstand- ing = 3.86; F (1, 75) = 3.39, p = .07). There was also a significant ef-

Fujita, Junko (2013), “Michelin for the Masses: Japan’s Preacher Kristopher J. and Andrew F. Hayes (2008), “Asymptotic Standing Restaurants Head for New York,” http:// and Resampling Strategies for Assessing and Comparing www.reuters.com/article/us-japan-restaurants-standing- Indirect Effects in Multiple Mediator Models,”Behavior idUSBRE94C04A20130513 Research Methods, 40 (3), 879-91. Labroo, Aparna A. and Jesper H. Nielsen (2010), “Half the Thrill is Strack, Fritz, Leonard L. Martin, and Sabine Stepper (1988), in the Chase: Twisted Inferences from Embodied Cognitions “Inhibiting and Facilitating Conditions of the Human Smile: and Brand Evaluations,” Journal of Consumer Research, 37 A Nonobtrusive Test of the Facial Feedback Hypothesis,” (1), 143-158. Journal of Personality and Social Psychology, 54 (5), 768-77. Lozano D.I., S.L. Crites and S.N. Aikman (1999), “Changes in Zellner, Debra A., Susan Loaiza, Zuleyma Gonzalez, Jaclyn Pita, Food Attitude as a Function of Hunger,” Appetite, 32 (2), Janira Morales, Deanna Pecora, and Amanda Wolf (2006), 207-18. “Food Selection Changes Under Stress,” Physiology & Pham, Michel, Iris W. Hung, and Gerald J. Gorn (2011), Behavior, 87 (4), 789-93. “Relaxation Increases Monetary Valuation,” Journal of Marketing Research, 48 (5), 814-26. Pizzagalli, Diego A., Ryan Bogdan, Kyle G. Ratner, and Allison L. Jahn (2007), “Increased Perceived Stress is Associated with Blunted Hedonic Capacity: Potential Implications for Depression Research,” Behavior Research and Therapy, 45 (11), 2742-53. When Less is More: Not Showing the Product’s Picture in an Advertisement Can Increase Ad and Brand Evaluations Tao Tao, Hong Kong Baptist University, Hong Kong, China Leilei Gao, Chinese University of Hong Kong, China

EXTENDED ABSTRACT mystery, enhancing favourableness. Second, consumers might try to Advertisers often include both the focal product’s picture and imagine the appearance of the product when it is absent (Sengupta a context picture that illustrates the brand’s benefits in a print ad- and Gorn 2002), and the self-generated product’s image might be vertisement. In some cases, a product’s visual appearance is an im- more attractive. Third, the product’s picture might be unattractive portant information to the viewers (Lazarus 1982, 1991; Yeung and and thus lead to negative impressions. To rule out these explana- Wyer 2004). In other cases, however, presenting the focal product’s tions, Experiment 2 manipulated whether the product’s picture was picture does not provide much additional information, especially revealed after participants had browsed an ad for tea but before they when the product mainly provides non-visual benefits (e.g., service), made evaluations. If the positive product omission effect was driven or when consumers are already familiar with the product’s appear- by any of the above explanations, this effect should be eliminated ance (e.g., coffee, tea). Then, when presenting the product’s picture after the product revelation. Results showed that ad evaluation was does not provide additional value to the ad viewers, how will the higher when product’s picture was not shown than when it was (4.64 absence (vs. presence) of a product’s picture affect consumers’ evalu- vs. 3.82; F(1, 115) = 9.37, p < .01), independent of product revela- ations of the ad and the band being advertised? The present research tion (Fs < 1). bears on this question. Experiment 3 and 4 tested an important moderator. We assume Previous literature has suggested that people respond to pictures that appreciation of the context picture requires holistic processing. in a holistic manner (Spoehr and Lehmkuhle 1982), which facilitates However, if the context picture is highly relevant to the viewers’ cur- the appreciation of the picture’s aesthetics and feelings (Holbrook and rent life experiences, they are likely to integrate his or her relevant Moore 1981). Although responses to pictures are typically holistic, a personal experiences into the simulation of the content, which in- visually prominent feature or an a priori motive to inspect a specific volves a detailed and concrete processing strategy (Zhao, Dahl and element can induce a shift from a globalized to a localized processing Hoeffler 2014). As a result, the global processing manner triggered manner (Spoehr and Lemhkuhle 1982; Hochstein and Ahissar 2002). by the absence of a product’s picture will impair this self-relevant Based on this logic, consumers are assumed to approach an ad in a imagery and decrease the affective reactions evoked by the simulated global manner. However, since consumers are normally aware of the experience. Consequently, omitting the product’s picture will have promotional intent of an advertisement, their subsequent attention is negative effects. likely to be quickly directed to the focal product’s picture (if any). To test this proposition, participants in Experiment 3 browsed As a result, consumers are likely to quickly shift from processing the an ad for a bridal cake. The context picture depicted a hugging cou- holistic features of the ad to attending to the product’s picture, facili- ple. Participants indicated their current romantic relationship status. tating a local processing manner. Furthermore, since the appreciation If they were in a relationship, their simulation was likely to involve of the context picture that conveys feelings often requires holistic their personal experiences, resulting in a negative product omission processing (Escalas 2014; Green and Brock 2000), a local process- effect. If participants were not, however, a positive product omission ing manner will reduce consumers’ appreciation of the affect-laden effect should occur. A significant interaction of product’s picture and benefits. Thus, we propose that when the product’s appearance can- relationship status (F(1, 117) = 6.47, p < .05) confirmed this predic- not contribute much to an appraisal of its quality, the omission of a tion. Participants who were not in relationship evaluated the brand picture of the product in an advertisement can increase consumers’ more favorably when the product’s picture was absent than when affective responses to the ad, and in turn can increase ad and brand it was present (6.76 vs. 6.22; F(1, 117) = 3.71, p =.06). In contrast, evaluations as well as purchase intentions. participants who were in a relationship rated the brand less favorably Using four different sets of stimuli, the first set of experiments in the former condition than in the latter condition (6.29 vs. 5.74; examined the positive effect of the omission of a product’s picture. F(1, 117) = 2.88, p =.09). Experiment 4 replicated these findings us- Experiment 1a employed a 2 (product presence: absent vs. present) x ing an ad of coffee and manipulating the content of the ad as either a 2 (ad stimulus: Lay’s potato chips vs. Turkish Airlines) between-sub- study or a business scenario, which was either relevant to the student jects design. The Lay’s chips ad depicted a scene of a girl enjoying participants or not. eating. The product’s picture, a bag of potato chips was either printed In summary, four experiments demonstrated that leaving out on the lower left-hand side of the ad or not. The Turkish Airlines ad (vs. presenting) the product’s picture in an ad can lead to a positive showed the silhouette of a girl standing on the beach, looking up effect on evaluations of the ad and the brand. These findings provide into the sky. The product’s picture, an airplane, was either presented practical implications to advertisement design. in the sky or absent from the ad. As expected, regardless of the ad stimulus, compared to the presence of a product’s picture, the ab- REFERENCE sence of the picture increased brand evaluation (6.35 vs. 7.02; F(1, Escalas, Jennifer Edson (2014), “Imagine Yourself in the Product: 114) = 4.76, p < .05), ad evaluation (5.02 vs. 5.51; F(1, 114) = 4.42, Mental Simulation, Narrative Transportation, and Persuasion,” p < .05), and affective reaction (4.46 vs. 5.09; F(1, 114) = 6.64, p < Journal of Advertising, 33(2): 37-48. .05). Besides, bootstrapping analyses showed that affective reaction Green, Melanie C., and Timothy C. Brock (2000), “The Role of mediates the effects of product presence on brand and ad evaluations. Transportation in the Persuasiveness of Public Narratives,” Experiment 1b and 1c replicated these findings using an ad of foot Journal of Personality and Social Psychology, 79(5): 701-21. massage and an ad of tea ware, respectively. Hochstein, Shaul, and Merav Ahissar (2002), “View from the Top: Experiment 2 tested three alternative explanations. First, con- Hierarchies and Reverse Hierarchies in the Visual System,” sumers might perceive the absence of the product’s picture as a Neuron, 36(5): 791-804.

Holbrook, Morris B., and William L. Moore (1981), “Feature Spoehr, Kathryn T., and Stephen W. Lehmkuhle (1982), Visual Interactions in Consumer Judgments of Verbal Versus Pictorial Information Processing. San Francisco, CA: Freeman Presentations,” Journal of Consumer Research, 8(1): 103-13. Company. Lazarus, Richard S (1982), “Thoughts on the Relations between Yeung, Catherine WM, and Robert S. Wyer (2004), “Affect, Emotion and Cognition,” American Psychologist, 37(9): 1019- Appraisal, and Consumer Judgment,” Journal of Consumer 24. Research, 31(2): 412-24. ______(1991), Emotion and Adaptation, New York: Oxford Zhao, Min, Steve Hoeffler, and Darren W. Dahl (2009), “The University Press. Role of Imagination-Focused Visualization on New Product Sengupta, Jaideep, and Gerald J. Gorn (2002), “Absence Makes Evaluation,” Journal of Marketing Research, 46(1): 46-55. the Mind Grow Sharper: Effects of Element Omission on Subsequent Recall,” Journal of Marketing Research, 39(2): 186-201. “I Am What You Eat”: Parents’ Morality is Inferred from the Products they Choose for their Children Stefanie M. Tignor, Northeastern University, USA Remi Trudel, Boston University, USA

EXTENDED ABSTRACT 1965). One particularly salient cue that consumers utilize to infer The current research utilizes and extends attribution theory behavioral consistency with respect to food consumption is body (Heider 1958; Kelley 1967) to develop a model of “parental sur- weight; consumers typically infer that individuals who are over- rogate consumption,” or, purchasing products or services that one’s weight must consistently consume calorie-laden foods (Brownell child will subsequently consume. Most broadly, we argue that pa- 1991). In the current research we examine moral attributions formed rental surrogate consumption represents a moral consumption be- based on single, one-time parental surrogate consumption behaviors, havior, as such consumption stands to directly impact the health and and manipulate perceptions of consistency using children’s physical well-being of another individual (i.e., the child) who does not pos- characteristics. sess regular or complete control over the goods or services he or she consumes (Vitell 2003). From this framework, we demonstrate that Hypothesis 2: Consumers utilize children’s physical charac- perceptions of parents’ morality are influenced by the consumption teristics as a proxy for consistency information, choices they make for their children, and show that the strength of and adjust their moral attributions accordingly. such moral attributions is dependent on perceptions of behavioral Control. The second key determinant for forming character at- consistency and control. tributions is perceptions of control. According to attribution theory, Conceptual Background character attributions are formed only when individuals are perceived to be making a volitional, intentional choice (Jones and Da- Parental Surrogate Consumption vis 1965). There exists a rich body of work examining how people Parents represent a sizeable market; in the United States alone form attributions of consumers based on their own consumption 53 percent of consumers aged 18 to 40 report having at least one decisions (e.g., Vartanian, Herman, and Polivy 2007; Williams and child (Newport and Wilke 2013). In the European Union, nearly 30 Steffel 2014). Yet, to our knowledge, researchers have not yet exam- percent of all households contain at least one child (Eurostat 2016). ined whether attributions are similarly formed based on the products These parents as a group spend billions annually on products and ser- consumers select for others. Attribution theory would suggest that vices for their children, from toys and sports leagues to food, cloth- in such instances character attributions will be formed for the indi- ing, and medical care (Lino 2014). Yet to date consumer researchers vidual in control of the consumption decision (the parent), but will have not examined the impact that routinely making such interde- not be extended to the eventual consumer (the child). Furthermore, pendent choices has on parental consumers. Rather, the bulk of ex- parents should be relieved of blame for any unhealthy consumption tant research on parent-child consumption has focused on consumer behaviors made by their children outside of their supervision. socialization (e.g., John 1999) or children’s perceptions of products or advertising (e.g., Zhang and Sood 2002). Given the powerful role Hypothesis 3a: Consumers do not form moral attributions of parents play in children’s consumption behavior, it is important for children based on parental surrogate consump- consumer researchers to examine how routinely making these in- tion information, as they are not perceived as be- terdependent choices—hereto named “parental surrogate consump- ing in control of their consumption choices. tion”—impacts parental consumers. Importantly, the non-elective and power imbalanced nature of parental surrogate consumption Hypothesis 3b: Consumers only form moral attributions of par- distinguishes it from previously examined forms of marketplace sur- ents based on parental surrogate consumption rogacy (Solomon 1986). Though parental surrogate consumption can information to the degree that parents are per- take many forms, in the current research we examine the consump- ceived as being in control. tion of food products, and hypothesize the following:

Hypothesis 1: Consumers view parental surrogate consump- Experiment 1: Establishing Parental Surrogate tion as a moral issue, and rate parents who choose healthy products for their children as Consumption as a Moral Consumption Behavior more moral than parents who choose unhealthy Experiment 1 aims to establish parental surrogate consumption products for their children. as a moral consumption behavior by demonstrating that consumers form moral attributions of parents based on the food products they select for their children. Attribution Theory When consumers make consumption decisions, they are con- Method veying information about themselves to onlookers (Belk 1988). At- Participants were 397 Amazon Mechanical Turk workers. Two tribution theory (Heider 1958; Jones and Davis 1965) examines how healthy and two unhealthy food items were selected from a pool of people utilize choice and behavioral information to form impressions ten pre-tested snacks: apples with peanut butter, yogurt and fruit, a of others’ character traits (i.e., form “internal” attributions; Weiner slice of cake, and a donut. All participants were randomly assigned to 1985). view one of these snack choices, then asked to rate its unhealthiness Consistency. According to attribution theory, character attribu- using two 6-point bipolar scales: healthy/unhealthy and wholesome/ tions are formed when behaviors are perceived as being consistent unwholesome (α = .92). with an individual’s typical pattern of behavior (Jones and Davis

After viewing each snack, participants were next presented with with his snack, participants were asked to rate the consistency with a hypothetical scenario stating the following: “8 year-old David has which they believed David’s mother selects that snack using five just gotten home from school, and it’s time for his after-school snack. items (e.g., “David’s mother usually selects snacks like this for Da- David’s mother gives him…” followed by their assigned snack. Par- vid to eat”; α = .94). Participants then provided moral judgments of ticipants then were asked to make ratings of David’s mother’s moral- David’s mother (α = .93). ity, using four 6-point bipolar scales adapted from Steim and Nemer- Results and Discussion off (1995): ethical/unethical, moral/immoral, kind-hearted/cruel, and Results were analyzed via moderated mediation analysis (Hayes caring/uncaring (α = .90), interspersed with four distractor scales. 2013; PROCESS Model 58). The first model regressed child weight, Results and Discussion snack choice, and the interaction between these two on perceptions Results of a one-way ANOVA confirmed a significant -differ of behavioral consistency, our mediator. The second model regressed ence in morality ratings between conditions (F(3, 393) = 64.90, p < child weight, perceptions of behavioral consistency, snack choice, .001, η2 = .33). Mothers who fed “David” apple slices (M = 5.23) and and the interaction between snack choice and consistency on moral yogurt and fruit (M = 5.21) were rated as the most moral; mothers judgments. For full results and variable codings. Importantly, the who fed “David” a donut (M = 3.98) and a slice of cake (M = 4.02) conditional indirect effect of child weight on morality was signifi- were rated as less moral. Across conditions, participants’ perceptions cant for both the yogurt condition (B = -.54, 95% CI = -.81 to -.31) of parents’ morality were negatively correlated with their percep- and the donut condition (B = -.80, 95% CI = -1.10 to -.52). Together, tions of snack unhealthiness (r(395) = -.60, p < .001), providing these results confirm H2 by identifying perceptions of behavioral support for H1. consistency as the mediating mechanism between child weight and moral perceptions, for both healthy and unhealthy products. Experiment 2: The Moderating Role of Children’s Appearance Experiment 4: The Role of Perceived Control Experiment 2 examined whether the strength of moral attribu- In experiment 4 we incorporate perceived control into our mod- tions, formed via parental surrogate consumption information, is el of parental surrogate consumption, by testing whether the choices moderated by children’s physical appearance. that parents make impact the moral judgments formed of the children who are subjected to such choices. Method Participants were 314 Amazon Mechanical Turk workers. Par- Method ticipants were randomly assigned to one of four conditions in a 2 Participants in experiment 4 were 315 adult Amazon Mechani- (child weight) x 2 (snack type) design. All participants in all condi- cal Turk workers. Participants were randomly assigned to one of four tions were presented with an image of “8-year old David.” In one conditions in a 2 (child weight) x 2 (snack type) design, nearly iden- condition David appeared to be of a healthy weight; in the second tical to that utilized in experiment 2. In contrast to prior experiments, condition David was overweight. All participants read the same sce- participants were asked to make moral judgments of David himself nario presented in experiment 1, with either a donut or fruit with (α = .90). yogurt as the snack. After reading this scenario, participants were Results and Discussion asked to make moral judgments of David’s mother (α = .92). The joint impact of child weight and food choice on perceptions Results and Discussion of David was tested using a 2 x 2 ANOVA. Results revealed a sig- The joint impact of child weight and snack choice on percep- nificant main effect of child weight, such that participants perceived tions of David’s mother was tested using a 2 x 2 ANOVA. Results overweight David to be less moral (M = 4.12) than healthy weight 2 revealed a significant main effect of child weight, such that partici- David (M = 4.52; F(1, 311) = 16.15, p <.001, ηp = .05). Importantly, pants perceived David’s mother to be less moral when her child was the main effect of food choice failed to reach significance (F(1, 311) 2 = 2.68, p = .10, η 2 = .01), confirming H3a. The interaction of food overweight (F(1, 310) = 58.09, p <.001, ηp = .16). The main effect p 2 choice and child weight also failed to reach significance (F(1, 311) of food choice was also significant (F(1, 310) = 155.37, p < .001, ηp 2 = .33), with David’s mother being perceived as more moral when she = .59, p = .44, ηp = .00). As David was not in control of his snack chose a healthy snack, once again confirming H1. The interaction choice, consumers did not form character judgments of him based on of food choice and child weight was significant (F(1, 310) = 12.57, his mother’s choices. 2 p < .001, ηp = .04), providing initial support for H2. Mothers who selected an unhealthy product for their child were punished more Experiment 5: The Joint Impact of Control and severely when that child was overweight. Consistency In experiment 5 we investigate the joint impact of perceptions Experiment 3: The Mediating Role of Consistency of control and consistency in forming moral attributions by manipu- This moderating role of child weight, as established in experi- lating control in our scenarios. ment 2, may be explained via the consistency dimension of attri- Method bution theory. When a mother selects an unhealthy product for her Participants in experiment 5 were 468 adult Amazon Mechani- overweight child, consumers likely assume that such a choice is re- cal Turk workers. Participants were randomly assigned to one of four flective of a consistent pattern of behavior. In experiment 3 we test conditions using a 2 (child weight) x 2 (locus of control) design. this mechanism directly. Participants were first shown overweight or healthy weight “David.” Method They then were presented with one of two scenarios, both of which Participants in experiment 3 were 405 adult Amazon Mechani- resulted in David consuming a donut as his snack. In the “mother cal Turk workers. Participants were randomly assigned to one of four control condition,” the scenario was identical to that utilized in pre- conditions using the same 2 (child weight) x 2 (snack type) design vious experiments. and scenarios employed in experiment 2. After viewing “David” 912 / “I Am What You Eat”: Parents’ Morality is Inferred from the Products they Choose for their Children

In the “David control condition,” participants read the follow- exist many other under-studied surrogate consumption contexts ing: “David just got home from school and it’s time for his after characterized by an unequal distribution of power or control (e.g., school snack. David’s mother left him yogurt with fruit in the re- consumption on the behalf of the elderly or disabled). Future re- frigerator to eat as a snack. David instead grabs a donut out of his search should examine these contexts in the interest of expanding backpack that he got from the school cafeteria, and eats that as his our knowledge of surrogate consumption, as well as in the interest snack.” This scenario was selected from a pool of six scenarios, pre- of addressing these under-served consumer populations (Pechmann tested for perceptions of control. Participants in all conditions rated et al. 2011). the consistency with which they believed David to consume donuts (α = .92). Finally, all participants were asked to make moral judg- REFERENCES ments of David’s mother (α = .89). Belk, Russell W. and Gregory S. Coon (1993), “Gift Giving as Agapic Love: An Alternative to the Exchange Paradigm Based Results and Discussion on Dating Experiences,” Journal of Consumer Research, 20 We tested our hypotheses using a moderated mediation analy- (December), 393-417. sis (Hayes 2013; PROCESS Model 14). The first model regressed Brownell, Kelly D. (1991), “Personal Responsibility and Control child weight on perceptions of behavioral consistency, our mediator. over our Bodies: When Expectation Exceeds Reality,” Health The second model regressed child weight, perceptions of behavioral Psychology, 10 (February), 303-10. consistency, locus of control, and the interaction between locus of Eurostat (2016), “Household Composition Statistics,” http:// control and consistency on moral judgments. For full results and ec.europa.eu/eurostat/statistics-explained/index.php/ variable codings. Household_composition_statistics. The conditional indirect effect of child weight on judgments of Hayes, Andrew F. (2013), An Introduction to Mediation, mothers’ morality was significant for the mother control condition Moderation, and Conditional Process Analysis: A Regression- (B = -.24, 95% CI = -.38 to -.12), but not for the child control condi- Based Approach, New York, NY: Guilford Press. tion (B = .06, 95% CI = -.10 to .21). These results support H3b by Heider, Fritz (1958), The Psychology of Interpersonal Relations, demonstrating that though a child’s weight provides observers with New York, NY: Wiley. information regarding the consistency with which he consumes un- Kelley, Harold H. (1967), “Attribution Theory in Social healthy products, the indirect effect of weight on moral judgments is Psychology,” Nebraska Symposium on Motivation. only significant for mothers who possess control. John, Deborah Roedder (1999), “Consumer Socialization of Children: A Retrospective Look At Twenty-Five Years of General Discussion Research,” Journal of Consumer Research, 26 (December), The current research extends attribution theory in establishing a 183-213. model of parental surrogate consumption. Across five experiments, Jones, Edward E. and Keith E. Davis (1965), “From Acts to we: 1) define and characterize parental surrogate consumption as a Dispositions: The Attribution Process in Person Perception,” moral consumption behavior, 2) demonstrate that just as moral attri- Advances in Experimental Social Psychology, 2, 219-66. butions are formed based on personal consumption information, so Komarova Loureiro, Yuliya, Julia Bayuk, Stefanie M. Tignor, too are they formed based on parental surrogate consumption infor- Gergana Y. Nenkov, Sara Baskentli, and David Webb (2016), mation, and 3) demonstrate that the occurrence and strength of moral “The Case for Moral Consumption: Examining and Expanding attributions is dictated by perceptions of behavioral consistency and the Domain of Moral Behavior to Promote Individual and control. Collective Well-being,” Journal of Public Policy & Marketing, Theoretical Implications 35 (November), 305-22. This research contributes to the consumer behavior literature in Lino, Mark. (2014). Expenditures on Children by Families, 2013. a number of ways. First, we extend attribution theory beyond judg- U.S. Department of Agriculture, Center for Nutrition Policy ments made based on the choices consumers make for themselves and Promotion. Miscellaneous Publication No. 1528-2013. (e.g., Olson et al. 2016; Williams and Steffel 2014). In this work, we Newport, Frank and Joy Wilke (2013), “Desire for Children Still demonstrate that such attributions also can be formed based on the Norm in U.S.,” http://www.gallup.com/poll/164618/desire- choices consumers make for others (specifically, their children). Sec- children-norm.aspx. ond, we direct attention to an overlooked yet highly impactful con- Olson, Jenny G., Brent McFerran, Andrea C. Morales, and sumption context: consumption on the behalf of children. Though Darren W. Dahl (2016), “Wealth and Welfare: Divergent considerable research has been conducted on child-relevant phe- Moral Reactions to Ethical Consumer Choices,” Journal of nomena such as consumer socialization (e.g., John 1999), the moral Consumer Research, 42 (April), 879-96. implications of the parental surrogate consumption process remain Pechmann, Cornelia, Elizabeth S. Moore, Alan R. Andreasen, largely unexplored. Third, we answer the call to “broaden the scope” Paul M. Connell, Dan Freeman, Meryl P. Gardner, Deborah of consumer marketplace morality by introducing parental surrogate Heisley, R. Craig Lefebvre, Dante M. Pirouz, and Robin L. consumption as a moral marketplace behavior (Komarova Loureiro Soster (2011), “Navigating the Central Tensions in Research et al. 2016). on at-risk Consumers: Challenges and Opportunities,” Journal of Public Policy & Marketing, 30, 23-30. Limitations and Future Research Solomon, Michael R. (1986), “The Missing Link: Surrogate Though the current research is the first of its kind, it is not with- Consumers in the Marketing Chain,” Journal of Marketing, 50 out limitations. Though we define parental surrogate consumption (October), 208-18. as pertaining to any number of products, herein we only investigate Steim, Richard I. and Carol J. Nemeroff (1995), “Moral Overtones food choices. Future research should investigate non-food products of Food: Judgments of Others Based on What They such as screen time, educational products, and medical care, among Eat,” Personality and Social Psychology Bulletin, 21 (May), others. Additionally, though we define parental surrogate consump- 480-90. tion as pertaining to the parent-child relationship specifically, there Advances in Consumer Research (Volume 45) / 913

Vartanian, Lenny R., Peter C Herman, and Janet Polivy (2007), Williams, Elanor F. and Mary Steffel (2014), “Double Standards in “Consumption Stereotypes and Impression Management: How the Use of Enhancing Products by Self and Others,” Journal of You Are What You Eat,” Appetite, 28, 265-77. Consumer Research, 41 (August), 506-25. Vitell, Scott J. (2003), “Consumer Ethics Research: Review, Zhang, Shi and Sanjay Sood (2002), “‘Deep’ and ‘Surface’ Cues: Synthesis and Suggestions for the Future,” Journal of Business Brand Extension Evaluations by Children and Adults,” Journal Ethics, 43 (March), 33-47. of Consumer Research, 29 (June), 129-41. Weiner, Bernard (1985), “An Attributional Theory of Achievement Motivation and Emotion,” Psychological Review, 92 (October), 548-73. The Implications of a Planned Development on Consumption Practices: A Case Study of an Orang Asli Community in Malaysia Ding Hooi Ting, Universiti Teknologi PETRONAS, Malaysia Mireya Sosa Abella, Universiti Malaya, Malaysia

EXTENDED ABSTRACT interpreted and compared against each other, following the stages This paper focuses on one of the Mah Meri’s community, an recommended by Charmaz and Belgrave (2012)— analysis and ini- Orang Asli community (Indigenous) in Peninsula Malaysia. The tial coding, analysis, and interpretation (focused coding), and lastly, Orang Asli communities in Malaysia are dispersed in many locali- the interpretation of the overarching themes were developed from ties all over Malaysia, and they are non-homogenous. The Mah Meri the focused codes. community in this case study occupied 36,000 acres of land during The themes that were derived from the analyses were material- their ancestral time (Endicott, 2015). The government intends to up- istic involvement, losing connection to nature, living sources, taboo grade the living quality of the community and in place, implemented beliefs, a daily habit, political system, new subsistence and beliefs a planned development on their land (their land was “exchanged” and future living. These themes were associated with the altered with a local agent who had altered the land into a plantation estate consumption practices that are mainly focusing on these broadly with the hope to provide employment and increased economic con- defined consumption themes: from a necessity consumption (basic, ditions of the community. With the intervention, the community has self-sustenance, subsistence, productive); and cultural consumption been relocated to places that are different from their original habitat, (self-referential, culture, reality) towards a higher order consump- of which the community needs to adapt to new conditions and envi- tion practices of materialistic consumption (functional, aesthetic, ronment. reflective, sensuous); playful consumption (performance, pleasure, Though the planned development by the government (to reduce entertainment, imaginary); intellectualized consumption (cognitive, poverty eradication and to create a balanced society) have their own signs, experience); trend consumption (fashion, statement, design) merits and benefits, these initiatives could be controversial for the and contingency consumption (incidental, emergency). Mah Meri community, whose spaces and natural environment are There is a gradual evolution of consumption practices from self- intruded. The planned development has modified the Mah Meri way sustenance consumption to a more materialistic-typed consumption of life and brought forth different consumption practices. The aim of practices (Spronk, 2014) with the introduction of the plantation es- this study is to analyze the consumption changes that the Mah Meri tate and small town. The presuppose consumers, still with minimal have had to adapt to due to the pressure of planned development— purchasing power, have contracted the habit of buying and liking from practices of self-sustenance and minimal market transactions to practical, functional products and consumer objects which provide a higher reliance on market transactions and material consumption. them with some form of consumption statement. This consumption For the Orang Asli, nature is the foundation of their subsistence pattern expresses a materialistic aspiration and ambition with the and culture (Munoz, Mladenoff, Schroeder, and Williams, 2014); change in the social structure (Warde, 2017). There is also a certain whereas for the contemporary world, the meaning of nature is con- reification of the mentality in the formation of consumption prac- nected mainly to development plans. During inception, the Mah Meri tices and how the boundaries of consumption are further upgraded community believe that the local agent (who are tasked to implement (Wahlen and Laamanen, 2015)—breaking away from the necessity the developments) will increase their welfare and economic condi- consumption purposes to a higher order consumption practices. The tions. But in reality, the development has created a different scenario monetarization of small town life becomes more prevalent. It be- to the community. The community is in conditions of the extreme comes evident that money quantifies qualitative differences, reduces marginal situation, which creates a low power defend for them. them to the monotony of more or less and the dialectic between the The alterations that have been brought by planned development poor and the rich is more visible and differentiated. have impacted and changed the lifestyle of the Mah Meri. Initially, In the case of Mah Meri community, they had tried to blend the Mah Meri has insisted that they would utilize their ancestral land, in their lifestyle by adapting to the new conditions. The govern- but the government believes that a planned development would pro- ment’s development plan has caused a trajectory in the lives of the vide far-reaching benefits on the community. With the land owner- Mah Meri and their connection to their natural spaces and cosmos. ship changed to the local agent, the majority of the community has The development has opened a new perspective and dimensions in rejected the work opportunities provided on the plantation. With the their daily lives. Achieving higher wealth and competitiveness are rise of the plantation estate and settlement (a small township has been certainly worthy pursuits in addressing the economic challenges established) and the loss of the natural spaces, the self-sustenance (Hayes, Caldwell, Licona, and Meyer, 2015). However, the condi- concept dwindles. For most of the Mah Meri who suffered the loss of tion set forth by the planned development has changed the lifestyle territories, they have to incorporate new ways of working and living. of the Mah Meri community. The findings showed supports that the This study used a phenomenology design to explore the Mah community’s views about living had been detached from the original Meri community. The researchers had collected data from the head view. The community has changed their consumption patterns from a of the Mah Meri community and 10 villagers in two different inter- situation of need and necessity to a relatively more aesthetic habitus view sessions that lasted for nine hours. The semi-structured probing practices when there are elements of development (Alba and Wil- questions were: Tell us about the life of an indigenous community liams, 2013; Pilar and Emontspool, 2015; Ulver and Ostberg (2014). (before and after the development); how the developmental changes are impacting the community and how the community is adapting REFERENCES to the changes. Interviews were recorded and transcribed. The tran- Alba, Joseph W. and Elanor F. Williams (2013), “Pleasure scripts were analyzed using thematic analysis, involving comparing Principles: A Review of Research on Hedonic Consumption,” transcripts for similarities and differences, and searching for the ma- Journal of Consumer Psychology, 23(1), 2-18. jor structural relationships that were related. The themes were then

Charmaz, Kathy and Liska Belgrave (2012), Qualitative Spronk, Rachel (2014), “Exploring the Middle Classes in Nairobi: Interviewing and Grounded Theory Analysis, The SAGE From Modes of Production to Modes of Sophistication,” Handbook of Interview Research: The Complexity of the African Studies Review, 57(1), 93-114. Craft, 2, 347-365. Ulver, Sofia, and Jacob Ostberg (2014), “Moving up, down or Endicott, Kirk (2015), Malaysia’s Original People: Past, Present sideways? Exploring consumer experience of identity and and Future of the Orang Asli, NUS Press. status incongruence,” European Journal of Marketing, 48(5/6), Hayes, Linda A., Cam Caldwell, Bryan Licona, and Thomas E. 833-853. Meyer (2015), “Followership Behaviors and Barriers to Wahlen, Stefan and Mikko Laamanen (2015), “Consumption, Wealth Creation,” Journal of Management Development, Lifestyle and Social Movements,” International Journal of 34(3), 270-285. Consumer Studies, 39(5), 397-403. Munoz, Samuel E., David J. Mladenoff, Sissel Schroeder, and Warde, Alan (2017), “The Development of the Sociology of John W. Williams (2014), “Defining the Spatial Patterns of Consumption,” in Consumption, UK: Palgrave Macmillan, Historical Land Use Associated with the Indigenous Societies 33-55. of Eastern North America,” Journal of Biogeography, 41(12), 2195-2210. Pilar, Rojas Gaviria, and Julie Emontspool (2015), “Global cities and cultural experimentation: cosmopolitan-local connections,” International Marketing Review, 32(2), 181- 199. Countdown or Countup: Effect of Direction of Time-keeping on Resource Deficiency and Related Downstream Preferences Sanjeev Tripathi, Indian Institute of Management, Indore, India Ankur Kapoor, Indian Institute of Management, Ahmedabad, India

EXTENDED ABSTRACT However, an alternative explanation is possible. Literature Time is a scarce resource (Caroll 2008) and individuals need to suggests that downward (upward) counting leads to higher (lower) accomplish tasks in a fixed time (deadlines, grocery shopping, exer- arousal, resulting in higher (lower) preferences for hedonic foods cise). Time-keeping can happen in two directions: countups, where like chocolate cake (Rook and Gardner 1993; Di Muro and Murray time is counted upwards (time-elapsed); or countdowns, where time 2012; Edith and Morwitz 2013). Study 2 tests this alternate explana- is counted downwards (time-left). While time-keeping is an impor- tion, by exploring the effect of time-keeping direction on risk prefer- tant and commonplace phenomenon, there is little research in this ences. While arousal leads to risk seeking (Horvath, and Zuckerman domain. This research aims to study the impact of direction of time- 1993; Mano 1994), resource deficiency can lead to risk-aversion keeping on food choices, risk preferences, and helping intentions; (March and Shapira 1992; Haushofer and Fehr 2014). Accordingly, and explain the effects through theoretical perspective of resource if downward time-keeping results in resource deficiency perception deficiency. Through four studies, we infer that downward (upward) (arousal), participants should exhibit risk-aversion (risk seeking be- time-keeping leads to more (less) resource deficiency, as manifested haviour). Participants (N=75) were randomly assigned to: upward/ in higher (lower) preference for calorie-rich food, more (less) risk- downward time-keeping (similar to Study 1) with risk preference aversion and lower (higher) helping intentions. (pre-tested polar scale with 1=50% chance of winning 200 units of Time is considered a resource in many decision making sce- local currency, and 11=sure amount of 90 units; lesser value means narios (Becker 1965; Leclerc, Schmitt, and Dube 1995; Festjens and higher risk preference) as dependent measure. As expected, partici-

Janiszewski 2015). This is supported by casual observation about in- pants in downward time-keeping had lower preference for risk (Mup- dividuals complaining about perpetual time shortage. The key prem- ward-time-keeping=5.37, Mdownward-time-keeping=6.88, t(69)=1.86,p=.032). These ise of this research is that the direction in which the consumption results support our earlier theorization that downward time-keeping of time is monitored influences the resource deficiency perception, leads to higher resource deficiency perception, and effects are not leading to downstream effects. We argue that downward (upward) driven by arousal. time-keeping leads to higher (lower) resource deficiency perception, While Study 2 rules out the alternate explanation of arousal, it because decreasing (increasing) sequences are implicitly associated is still possible that the effects of downward time-keeping are due with reducing (increasing) quantities, value or resources (Pelham, to presence of zero as hard stop. Zero symbolizes completion and Sumarta, and Myaskovsky 1994; Pandelaere, Briers, and Lembregts exhaustion (of resources). It is possible that resource deficiency per- 2011). There is general association of decreasing sequences with ception is induced by presence of zero as the end-point in down- scarcity and reduction (reduction in stock prices, bank balance, calo- ward time-keeping, not because of the decreasing sequence. Study rie consumption etc.). Therefore, a decreasing sequence (especially 3 investigates this alternate explanation. If the phenomenon is due for resource like time) signals resource reduction more strongly to presence of zero as hard stop, the effects found in earlier stud- than an increasing sequence. Accordingly, we argue that downward ies should not be observed in downward time-keeping with non-zero (upward) time-keeping (60,59,…vs…59,60) is more (less) likely to ending, but if it is due to decreasing sequence, this effects should be induce resource deficiency perception. Further, such resource de- replicated. Participants (N=60) were randomly assigned to: upward ficiency has been found to be associated with higher self-focused (5-64 sec) or downward (64-5 sec) time-keeping (rest of the design behaviors (Levontin, Ein-Gar and Lee 2014), consumption of cal- was identical to Study 2). It was found that participants in downward orie-rich foods (Briers and Laporte 2013) and risk-aversion (March time-keeping had lower preference for risk (Mupward-time-keeping=6.68, and Shapira 1992; Haushofer and Fehr 2014). Hence downward (up- Mdownward-time-keeping=8.09,t(46)=1.84,p=.035) than in upward time- ward) time-keeping should lead to more (less) resource deficiency, as keeping. The results suggest that it is not the end point (zero or manifested in higher (lower) preference for calorie rich food, more non-zero) but the decreasing or increasing sequence of time-keeping (less) risk-aversion and less (more) willingness to help. that induces resource deficiency perception and related downstream In Study 1, participants (N=95) were randomly assigned to effects. 2(Downward/Upward time-keeping)x2(Evaluation of: Calorie-rich/ Finally, Study 4 tests the mediation of perceived resource defi- Calorie-lean food) between-subjects conditions. They worked on a ciency, and attempts to increase the validity of the results with dif- task (crossing e’s) timed either in upward (0-60 seconds) or down- ferent operationalization of timed task. Participants (N=132) were ward (60-0 seconds) direction. Next, they evaluated either a calo- randomly assigned to: upward/ downward time-keeping with task of rie-rich (chocolate cake) or a calorie-lean (fruit salad) food item on finding words in a word-block, while the time was ticking on the side purchase likelihood (Dodds, Monroe, and Grewal 1993). Two way of screen. Next, they provided responses for willingness-to-donate ANOVA revealed a significant interaction between time-keeping money to and willingness-to-help a charitable organization, followed direction and food category (F(1,89)=7.195,p=.009). Planned con- by response to a 5-item perceived resource deficiency scale adapted trasts revealed a significant difference in preference for calorie- from Levontin, Ein-gar and Lee (2014). It was found that partici- rich food between upward and downward time-keeping conditions pants in downward (upward) time-keeping had less (more) helping intentions (M =7.25, M =6.73,t(128)=2 (Mupward-time-keeping=6.97, Mdownward-time-keeping=8.02,t(44)=2.65,p=.005) but upward_time-keeping downward_time-keeping non-significant difference for calorie-lean food item. The results are .27,p=.012). Mediation analysis (Hayes 2013; Model 4) supported consistent with the possibility that downward (upward) time-keeping mediation of perceived resource deficiency between time-keeping leads to higher (lower) resource deficiency, as manifested in higher direction and willingness to help (95%CI for indirect effect: 0.0158- (lower) preference for calorie-rich food. 0.3089). We also measured affect (PANAS; Watson, Clark, and Tel- legen 1988), and found no difference in different time-keeping con-

Advances in Consumer Research 916 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 917 ditions. This helps us rule out mood or stress based explanation of Horvath, Paula, and Marvin Zuckerman (1993), “Sensation seeking, results. risk appraisal, and risky behavior.” Personality and Individual This research is among the first to study the direction of time- Differences, 14(1), 41-52. keeping and we hope that it provides direction for richer understand- Leclerc, France, Bernd H. Schmitt, and Laurette Dube ing of time perception. (1995), “Waiting time and decision making: Is time like money?.” Journal of Consumer Research, 22(1), 110-119. REFERENCES Levontin, Liat, Danit Ein-Gar, and Angela Y. Lee (2015), “Acts of Caroll, Joseph. (2008), “Time Pressures, Stress Common for emptying promote self-focus: A perceived resource deficiency Americans”, Retrieved from: http://www.gallup.com/ perspective.” Journal of Consumer Psychology, 2(25), 257- poll/103456/time-pressures-stress-common-americans.aspx 267. Becker, Gary S. (1965), “A Theory of the Allocation of Time.” The Mano, Haim (1994), “Risk-taking, framing effects, and Economic Journal, 75(299), 493-517. affect.” Organizational behavior and Human Decision Briers, Barbara, and Sandra Laporte (2013), “A wallet full of Processes, 57(1), 38-58. calories: The effect of financial dissatisfaction on the desire for March, James G., and Zur Shapira (1992), “Variable risk food energy.” Journal of Marketing Research, 50(6), 767-781. preferences and the focus of attention.” Psychological Di Muro, Fabrizio, and Kyle B. Murray (2012), “An arousal review, 99(1), 172-183. regulation explanation of mood effects on consumer choice.” Pandelaere, Mario, Barbara Briers, and Christophe Lembregts Journal of Consumer Research, 39(3), 574-584. (2011), “How to make a 29% increase look bigger: The Dodds, William B., Kent B. Monroe, and Dhruv Grewal (1991), unit effect in option comparisons.” Journal of Consumer “Effects of price, brand, and store information on buyers’ Research, 38(2), 308-322. product evaluations.” Journal of Marketing Research, 28(3), Pelham, Brett W., Tin T. Sumarta, and Laura Myaskovsky 307-319. (1994), “The easy path from many to much: The numerosity Edith, Shalev, and Vicki Morwitz (2013), “Does Time Fly When heuristic.” Cognitive Psychology, 26(2), 103-133. You’re Counting Down? The Effect of Counting Direction Rook, Dennis W., and Meryl P. Gardner (1993), “In the mood: on Subjective Time Judgement.” Journal of Consumer Impulse buying’s affective antecedents.” Research in Psychology, 23(2), 220-27. Consumer Behavior, 6(7), 1-28. Festjens, Anouk, and Chris Janiszewski (2015), “The value of Watson, David, Lee A. Clark, and Auke Tellegen (1988), time.” Journal of Consumer Research, 42(2), 178-195. “Development and validation of brief measures of positive and Haushofer, Johannes, and Ernst Fehr (2014), “On the psychology of negative affect: the PANAS scales,” Journal of Personality poverty.” Science, 344(6186), 862-867. and Social Psychology, 54 (6), 1063-1070. Hayes, Andrew F (2013), Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York: Guilford Press. How Incidental Confidence Influences Self-Interested Behaviors? A Double-Edged Sword Claire Tsai, University of Toronto, Canada Jia lin Xie, University of Toronto, Canada

EXTENDED ABSTRACT of others, people primed with status-enhancement motives are more When reading the daily news, one finds examples too numer- likely to purchase costly green products to signal their willingness ous to count of politicians, sports stars, business executives, students and ability to bear the cost of prosocial consumption (Griskevicius from top schools, and others finding all sorts of ways to profit by et al. 2010). However, this preference reverses in a private shop- behaving unethically (e.g., Netter 2010). The current research adds to ping context (e.g., online shopping) because prosocial consumption the literature on self-interested behaviors by investigating a hitherto would have little signaling value. unexplored, yet ubiquitous, determinant—incidental confidence. In summary, we propose that lower incidental confidence in- People make decisions and take actions based not only on what creases (decreases) self-interested behaviors when money (altruism) they like or believe, but also on their confidence in their likes and be- is the primary signal of status that helps compensating for confi- liefs. For example, lower confidence can decrease charitable giving dence. We conducted four experiments to test these hypotheses. (Tsai and McGill 2011), purchases (Greenleaf and Lehmann 1995), price expectations (Mazumdar and Jun 1993; Urbany and Dickson Study 1

1991), the amount of money bet on sports games (Tsai, Klayman, Method. Study 1 (Ntarget = 150; Nactual = 121; 36.60% female; age and Hastie 2007), and willingness to engage in competition (Camerer = 34.21; Amazon Mechanical Turk) tested the hypothesis that lower and Lovallo 1999). incidental confidence would increase self-interested behaviors relat- However, most studies of consequences of confidence have ing to money using a one-shot dictator game, a task commonly used examined how characteristics of a focal event influence confidence to demonstrate self-interested behaviors involving monetary conse- in judgments of the same event and subsequent decisions about the quences (Camerer 2003). We had a target sample size of 75 partici- event (Yates, 1990). It is curious whether changes in confidence in pants per cell and managed to recruit 142 participants. Twenty-one one domain can influence behaviors in other unrelated domains. We participants failed to pass an attention check and were excluded from add to this line of work by investigating such transfer effect of in- the analysis, resulting in 121 valid data points. cidental confidence on self-interested behaviors. Self-interested be- Study 1 employed a two-level single-factor (incidental confi- haviors are considered by many researchers a balancing act between dence: lower vs. higher) between-subjects design. We first manipu- wanting to do good and wanting to profit from doing bad (Eisenberg lated incidental confidence by asking participants to list five events and Shell 1986). The present research examines how incidental con- in which they experienced high or low confidence in their judgments fidence might tilt the balance and increase selfish or selfless behav- or ability, a procedure adapted from existing research (Petty et al., iors. 2002). Pretesting (N = 46) ruled out power and self-esteem as poten- In the absence of a subjective sense of confidence, people often tial confounds. experience a state of psychological aversion that they are motivated Next, participants were endowed with $3.00 and allocated the to reduce (Kahneman, Slovic, and Tversky 1982; Payne, Bettman, fund with an unknown recipient in a one-shot anonymous dictator and Johnson 1992). When it is not possible to restore confidence by game. Participants kept whatever he or she did not offer; the recipient improving judgments or ability in the same domain, people can re- could choose to accept or reject the offer. store confidence by gaining status or increasing rankings in unrelated Results. As predicted, participants in the lower-confidence con- domains (Coon 1994). For example, an individual who writes poorly dition behaved more selfishly; they kept more money for themselves may regain his/her confidence by excelling in mathematics. In the than participants in the high-confidence condition (Mlow-conf = $2.31, context of our study, the desire to compensate for lower confidence SD = .64 vs. Mhi-conf = $2.04, SD = .72; t(119) = –2.19, p = .03; d = may cause people to seek status. The status motives, in turn, can .40). increase self-interested behaviors that generate financial gains or encourage selfless behaviors, depending on whether money or altruism, Study 2 respectively, is a more prominent source of status. Method. Study 2 (Ntarget = 100; Nactual = 80; 56.10% female; age On the one hand, money is a prominent resource and can serve = 19.52; university students) expanded the research to another self- as an important signal of status (Kemper 1991). Money brings about interested behavior: prosocial consumption that involved a tradeoff a state of self-sufficiency, provides a feeling of confidence that prob- between money (cost of product) and altruism. lems can be solved and needs can be met (Vohs, Mead, and Goode Participants first completed the confidence manipulation task 2006), and increases the confidence that one can obtain positive out- from Study 1. They then viewed an advertisement and indicated their comes and avoid negative ones (Johnson and Krueger, 2006). We preference (1 = non-green product, 7 = green product). The adver- therefore propose that the motive to compensate for lower confidence tisement featured a cheaper non-green product and a more-expensive can trigger self-interested behaviors that generate financial payoffs. green product. All the other product attributes were identical. On the other hand, lower incidental confidence can reduce self- To address the potential issue concerning negative affect, we interested behaviors when altruism is a stronger signal of status than measured mood and anxiety before and after the choice and con- money (Roberts, 1998). Although status can be achieved by force, trolled for them in the analyses. To rule out the possibility that some we focus on status that is achieved through prestige, freely conferred form of mood regulation might be involved, we measured anticipated deference (Henrich and Gil-White 2001). For example, public dis- happiness, “to what extent do you think having the product of your plays of self-sacrifice for strangers in the group can increase the self- choice would make you feel happy (7 = extremely). sacrificer’s status and the likelihood of being selected as a leader in Results. Consistent with Study 1, lower confidence reduced pro- that group (Gurven et al. 2000). Similarly, when shopping in front social consumption (Mlow-conf = 4.38, SD = 1.31 vs. Mhi-conf = 3.70, SD

= 1.63; t(78) = –2.03, p = .04; d = .46). (Higher number indicated the savings incurred from purchasing a non-green product was con- weaker preference for the green product.) Mood, anxiety, and antici- sidered more effective in compensating for lower confidence. These pated happiness did not vary across conditions (all p > .30). When results were consistent with the proposed compensatory mechanism. we controlled for these variables in separate ANCOVAs, none of The results further ruled out negative affect, perceived fairness, and them had an effect on product preferences (allp > .33), and the effect effort as alternative explanations because these alternative explana- of confidence remained significantall ( p = .04). tions implied a main effect of confidence regardless of the source Discussion. The results of Studies 1 and 2 support our hypoth- of status. Instead, we found that the effect of confidence on product esis that lower confidence can increase self-interested behaviors that preference was not significant (F < 1, p = .59), but the two-way in- generate monetary payoffs. These findings show that confidence can teraction of confidence and source of status was. have a transfer effect on subsequent self-interested behaviors that are unrelated to the true source of confidence. Study 4 In the next two experiments, we investigated the underlying In Study 4, we further tested the compensatory mechanism process by manipulating source of status directly—money versus using the same one-shot dictator game from Study 1 and another altruism. Our proposed compensatory mechanism implies that lower procedure to manipulate source of status. Specifically, we varied the confidence would decrease prosocial consumption (Study 3) and in- strength of group affiliation (Briley and Wyer 2002). Prior research crease the amount of money kept for self in a dictator game (Study shows that sharing similar experiences can cause people to iden- 4) when money serves as the primary means for compensating for tify with the reference group (Aronson, Wilson, and Akert 2010). lower confidence. However, the results should reverse when altruism We therefore predicted that group affiliation would be strengthened can better compensate for lower confidence than money. when participants thought their recipients took the same study. As mentioned earlier, the display of altruism could compensate Study 3 for lower confidence because selfless behaviors could help one gain We expanded the research by manipulating source of status di- status. Moreover, prior studies on competitive altruism and costly rectly and using another procedure to manipulate incidental confi- signaling (Barclay and Willer 2007; Roberts 1998; Van Vugt et al. dence. The task was an abstract elemental task (Raven’s puzzles) that 2007) suggested that altruism was more likely to help one gain status further ruled out power, self-esteem, and even self-efficacy because in a community with strong group affiliation than one with weak solving puzzles is unlikely to alter social relations or control over group affiliation. Thus, altruism should be perceived as the better valued resources (power; Magee and Galinsky, 2008), self-accep- means for restoring confidence when group affiliation was strength- tance or worthiness (self-esteem; Wylie, 1979), and general belief ened and thus lower confidence should decrease the amount of in one’s own ability to complete tasks and reach goals (self-efficacy; money that people kept for themselves, causing them to appear more Bandura, 1997). altruistic. However, under the condition of weak group affiliation,

Method. The study (Ntarget = 120; Nactual = 115; 56.10% female; money would be the better means for restoring confidence. Thus, age = 19.52; university students) employed a 2 (confidence: lower lower confidence would increase the amount that people kept for vs. control) x 2 (source of status: money vs. altruism) between-sub- themselves, causing them to appear more selfish. jects design. We first manipulated incidental confidence by asking Method. The study (Ntarget = 120; Nactual = 103 university stu- participants to complete a set of 12 Ravens puzzles that were ei- dents) employed a 2 (confidence: lower vs. control) x 2 (source of ther difficult or easy. Pretesting (N = 37) confirmed that the puzzle status: money vs. altruism) between-subjects design. Participants task altered confidence level as intended (Mlow-conf = 4.06 vs. Mhi-conf = received $5 each for participation and 12 quarters at the beginning 8.80; t(35) = 6.61, p < .01). of the study. Participants first completed a puzzle task that primed Next, participants encountered the manipulation of source of incidental confidence as Study 3. status. Specifically, we told half the participants to consider shopping Next, participants proceeded to the one-shot dictator game from online by themselves (a private context) or in an actual store (a pub- Study 1 and encountered the manipulation of source of status. Spe- lic context; Griskevicius et al. 2010). Next, participants completed cifically, we told participants that their recipients either completed the shopping task from Study 2. the same puzzle study as they did (strong affiliation) or did a different We expected prosocial consumption to be a stronger signal of study (weak affiliation). They then divided their three-dollar endow- status than money when people considerer shopping in a public con- ment with their recipients. text (e.g., in a brick-and-mortar store). The notion of competitive al- Pretesting (N = 42 students) confirmed that lower confidence truism (Roberts 1998) suggests that people are particularly sensitive increased the motives for obtaining status-enhancing monetary re- to the social and reputational aspects of altruism (Bateson, Nettle, sources and that our confidence manipulation task did not induce and Roberts 2006). Therefore people are more likely to engage in negative affect or change perceived fairness of the study. The results prosocial consumption when shopping in public than in private. showed that participants in the low-confidence condition were more Results. A two-way ANOVA revealed a significant interaction inclined to attribute their desire for the monetary prize to status-en- of confidence and source of status only (F(1, 111) = 9.40, p = .002, hancement reasons (M = 3.43, SD = 1.57) than did participants in η2 = .075). Planned contrasts showed that when people considered the high-confidence condition (M = 2.39, SD = 1.39; t(42) = 2.32, shopping in private, lower confidence reduced prosocial consump- p = .02). Positive mood, anxiety, and fairness did not differ across tion (Mlow-conf = 4.81, SD = 1.82 vs. Mhi-conf = 3.82, SD = 2.14; F(1, conditions (all p > .20). 111) = 4.11, p = .04, d = .50), replicating the results of Study 2. How- Results and Discussion. A two-way ANOVA revealed a signifi- ever, the effect reversed when people considered shopping in public cant interaction of confidence and source of status only (F(1, 99)

(Mlow-conf = 4.52, SD = 2.02, vs. Mhi-conf = 5.69, SD = 1.46; F(1, 111) = = 6.15, p = .005, η2 = .076). Planned contrasts showed that when 5.30, p = .02, d = .66). participants thought that their recipients had taken a different study Discussion. The results showed that individuals primed with (weak affiliation, money as the source of status), lower confidence lower confidence appeared to be more or less prosocial in product increased the amount of money that they kept (Mlow-conf = $2.42, SD choices, depending on whether the ownership of a green product or = .64 vs. Mhi-conf = $2.01, SD = .72; F(1, 99) = 4.29, p = .04, d = 920 / How Incidental Confidence Influences Self-Interested Behaviors?A Double-Edged Sword

.60). However, the effect reversed when participants thought their French, John R. Jr. and Bertram Raven (1959), “The Bases of recipients had taken the same study (strong affiliation, altruism as the Social Power,” in Studies in Social Power, ed. Dorwin source of status; Mlow-conf = $1.86, SD = 1.07 vs. Mhi-conf = $2.37, SD = Cartwright, Ann Arbor, MI: Institute for Social Research, .72; F(1, 99) = 3.60, p = .04, d = .56). These results further supported 150–67. our proposed compensatory mechanism. Greenleaf, Eric A. and Donald R. Lehmann (1995), “Reasons for Substantial Delay in Consumer Decision Making,” Journal of General Discussion Consumer Research, 22 (2), 186–99. It is imperative to better understand the psychology of self- Griskevicius, Vladus, Joshua M. Tybur, and Bram Van den Bergh interested behaviors because they are often consequential and incur (2010), ”Going Green to be Seen: Status, Reputation, and high cost to the society. Our findings contribute to the confidence Conspicuous Conservation,” Journal of Personality and Social literature by documenting a transfer effect of incidental confidence Psychology, 98 (3), 392–404. on self-interested behaviors. Our work provides a clearer picture for Gurven, M., Allen-Arave, W., Hill, K., & Hurtado, M. (2000). It’s a how confidence influences self-interested behavior. In a series of wonderful life. Evolution and Human Behavior, 21, 263–282. four experiments, we found that people in a lower confidence state Henrich, J. & Gil-White, F. J. (2001). The evolution of prestige: may appear more selfish or selfless depending on the means for re- Freely conferred deference as a mechanism for enhancing storing confidence. Further, we tested our proposed compensatory the benefits of cultural transmission. Evolution and human mechanism in two studies by manipulating source of status in two behavior, 22(3), 165-196. different ways—shopping context (Study 3) and group affiliation Johnson, W., & Krueger, R.F. (2006). How money buys happiness: (Study 4). Finally, data ruled out alternative explanations including Genetic and environmental processes linking finances and life affect, fairness, effort, self-efficacy, power, and self-esteem. satisfaction. Journal of Personality and Social Psychology, 90, This research has important implications for the general area of 680–691. studies on ethical decision making. For unethical behaviors involv- Kahneman, Daniel, Paul Slovic, and Amos Tversky (1982), ing monetary payoffs (e.g., theft, fraud), incidental confidence may Judgment Under Uncertainty: Heuristics and Biases, New produce similar effects to those observed in the current research. In York, NY: Cambridge University Press. fact, we found in a post-test (N = 36 university students) that lower Kemper, Theodore D. (1991), “Predicting Emotions from Social confidence increased theft, a real transgression. Relations,” Social Psychological Quarterly, 54 (4), 330–42. It would be fruitful to further investigate the interplay between Mazumdar, Tridib and Sung Y. Jun, (1993), “Consumer Evaluations confidence and ethical decision making. Other promising avenues for of Multiple Versus Single Price Change,” Journal of Consumer future research include examining how the numerous variables af- Research, 20 (3), 441–50. fecting confidence (e.g., processing fluency; Tsai and Thomas 2011) Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The and source of status (e.g., expertise, respect from others; French and self‐reinforcing nature of power and status. The Academy of Raven, 1959) interact to influence self-interested behaviors. Management Annals, 2, 351–398. Netter, Sarah (2010), Confessions of a Ghostwriter: Man’s Career REFERENCES Thrives Helping Students Cheat. ABC News. Aronson, E., Wilson, T. D., & Akert, R. M. (2010). Social Payne, J. W., Bettman, J. R., & Johnson, E. J. (1992). Behavioral psychology. Upper Saddle River, NJ: Prentice Hall. decision research: A constructive processing perspective. Barclay, P., & Willer, R. (2007). Partner choice creates competitive Annual Review of Psychology, 43, 87–131. altruism in humans. Proceedings of the Royal Society of Petty, Richard E., Pablo Brinol, and Zakary L. Tormala (2002), London B: Biological Sciences, 274(1610), 749-753. “Thought Confidence as a Determinant of Persuasion: The Bandura, A. (1997). Self-efficacy: the exercise of control. New Self-Validation Hypothesis,” Journal of Personality and Social York, NJ: Freeman. Psychology, 82 (5), 722–41. Briley, Donnel and Robert R. Wyer (2002), “The Effects of Group Roberts, Gilbert (1998), “Competitive Altruism: From Reciprocity Membership on the Avoidance of Negative Outcomes: to the Handicap Principle,” Proceedings of the Royal Society Implications for Social and Consumer Decisions,” Journal of of London, Series B, 265, 427–431. Consumer Research, 29 (December), 400-15. Tsai, Claire I., Joshua Klayman, and Reid Hastie (2008), “Effects Camerer, Colin F. (2003), Behavioral Game Theory: Experiments of Amount of Information on Judgment Accuracy and on Strategic Interaction, Princeton, NJ: Princeton University Confidence,” Organizational Behavior and Human Decision Press. Processes, 107 (2), 97–105. Camerer, Colin F. and Dan Lovallo (1999), “Overconfidence Tsai, Claire I. and Ann L. McGill (2011), “No Pain, No Gain? How and Excess Entry: An Experimental Approach,” American Construal Level and Fluency Affect Consumer Confidence,” Economic Review, 89 (1), 306–318. Journal of Consumer Research, 37 (5), 807–21. Chambers, John R. and Paul D. Windschitl (2004), “Biases in Tsai, Claire I. and Manoj Thomas (2011), “When Does Feeling Social Comparative Judgments: The Role of Nonmotivated of Fluency Matter? How Abstract and Concrete Thinking Factors in Above-Average and Comparative-Optimism Effects, Influence Fluency Effects,”Psychological Science, 22 (3), Psychological Bulletin, 130 (5), 813–838. 348–54. Coon, Dennis (1994), Essentials of Psychology: Exploration Urbany, Joel E. and Peter R. Dickson (1991), “Consumer Normal and Application (6th ed.), St. Paul/Minneapolis, MN: West Price Estimation: Market Versus Personal Standards,” Journal Publishing Company. of Consumer Research, 18 (1), 45–51. Eisenberg, Nancy and Rita Shell (1986), “Prosocial Moral Judgment and Behavior in Children,” Personality and Social Psychology Bulletin, 12 (4), 426–33. Advances in Consumer Research (Volume 45) / 921

Van Vugt, M., Roberts, G., & Hardy, C. (2007). Competitive Wylie, R.C. (1979). The self-concept (Vol. 2). Lincoln: University altruism: development of reputation-based cooperation in of Nebraska Press. groups. In L. Barrett, & R. Dunbar (Eds.), Handbook of Yates, J.F. (1990). Judgment and decision making. Englewood evolutionary psychology (pp. 531-540). Oxford, England: Cliffs, NJ: Prentice-Hall. Oxford University Press. Vohs, K. D., Mead, N. L., & Goode, M. R. (2006). The psychological consequences of money. Science, 314, 1154- 1156. How Accepting Hierarchy Impacts Information Processing Lingjiang Lora Tu, Baylor University, USA Yinlong Zhang, University of Texas at San Antonio, USA

EXTENDED ABSTRACT Study 2, 3 and 4 provide convergent evidence for the proposed The current trend of marketers trying desperately to connect mechanism of consumption motive underlying the PDB effect. The with consumers emotionally, the so-called “sadvertising” (Fera bootstrapping analysis in study 2 supports the proposed mediation 2014) makes Aristotle’s (1354/1954) question new again: do people of consumption motive statistically. Mediation analysis was tested follow their hearts or heads? In this research, we aim to advance separately for affective and cognitive ads with Hayes (2013) Model understanding on this important consumer decision. 4 (with IV= PDB, Mediator = consumption motive, DV = Product Persuasive information can be put into two categories: affective Evaluation). The confidence intervals of the indirect effect of PDB information (i.e. information appealing to the heart such as feelings on product evaluation through consumption motive didn’t include and emotions) and cognitive information (i.e. information appealing zero for both affective and cognitive ads (affective ad: indirect effect to the head such as attributes and beliefs of the object) (See, Petty = -.12, Boot SE = .08, CI’s: -.33, -.01; cognitive ad: indirect effect and Fabrigar, 2008). To answer the call for more information pro- = .12, Boot SE = .07, CI’s: .02, .29), indicating consumption motive cessing research focusing on the global consumer and richer classi- mediates the effect of PDB on product evaluation for both the affec- fications of cultural differences (Johar, Maheswaran, and Peracchio, tive and cognitive ads. By manipulating consumption motive, study 2006), we identify Power Distance Belief (PDB hereafter) – the ex- 3 further confirms the causal path between the proposed mediator tent to which individuals expect and accept hierarchy and inequality and the dependent variable. Study 4 examines a boundary condition in society -- as a new factor that impacts the reliance of affective and of the PDB effect by manipulating external monitoring. As expected, cognitive information. More specifically, we hypothesize that con- the PDB effect manifests when external monitoring is non-salient sumers with low PDB rely more on affective information and prefer and attenuates when external monitoring is salient. With actual and affective ads whereas consumers with high PDB rely more on cogni- fictitious ads, across different product categories including products tive information and prefer cognitive ads. We further postulate that and services, and with both student and non-student participants, the this PDB effect is mediated by consumption motive with low PDB converging results demonstrate the robustness of the PDB effect on inducing consummatory motive whereas high PDB inducing instru- the persuasiveness of affective and cognitive information and sup- mental motive. Given the acceptance of one’s position in a hierarchy port the proposed process. Moreover, study 1a, 1b, and study 2 re- as the core of PDB, high PDB requires individuals to accommodate peatedly show that the PDB prime doesn’t impact situational self- to socially conferred positions in a hierarchy. The external regulation construal. Study 2 demonstrates that the PDB effect is not mediated of social judgments induces an external perceived locus of causality through self-construal. The experimental findings are consistent with (Heider, 1958; Botti and McGill, 2011) and promotes instrumental recent PDB research (Gao et al., 2016; Lalwani and Forcum, 2016) motive for high PDB consumers’ behaviors (Ryan and Deci, 2000). that PDB and self-construal are orthogonal theoretical constructs and The instrumental motive then induces a preference for products pre- exert impacts on consumer behavior via different mechanisms. sented with cognitive information (Dhar and Werternbroch, 2000). Our research makes several significant contributions. First, our In contrast, low PDB releases individuals from external regulation research enriches the growing body of evidence on consumers’ mo- of social judgments and promotes a more internal perceived locus tivated reasoning during information processing by identifying PDB of causality, which consequently induces consummatory motive for as a new factor that influences consumers’ selective preference for consumption behaviors (Deci and Ryan, 1985). The consummatory cognitive versus affective information. Second, this research broad- motive drives a preference for products presented with affective in- ens the cross-cultural persuasion research beyond the heavy reliance formation (Dhar and Werternbroch, 2000). on the INDCOLL framework (i.e. individualism-collectivism at the We propose external monitoring as a moderator for the PDB cultural level and independent-interdependent self-construal at the effect based on the classic motivation research findings that external individual level). By including PDB, this research greatly enhances regulation undermines consummatory motive and enhances instru- the predictive accuracy of cultural orientation on persuasive effec- mental motive (Deci, Koestner, and Ryan, 1999; Fishbach and Choi, tiveness. Third, this research contributes to the PDB literature (Gao, 2012; Maimaran and Fishbach, 2014). More specifically, we predict Winterich, and Zhang, 2016; Lalwani and Forcum, 2016; Winterich that salient external monitoring tends to regulate low PDB individu- and Zhang, 2014; Zhang et al., 2010) by uncovering a novel link be- als’ behavior by enhancing their instrumental motive. In contrast, tween PDB and consumers’ processing of affective versus cognitive because high PDB individuals’ baseline tendency to regulate their information and identify the underlying mechanism and a boundary behaviors with external regulation is already high (Friesen et al., condition. 2014), salient external monitoring should have little effect on high PDB individuals’ motive. REFERENCES Across one content analysis and five experimental studies, we Aristotle (1954), Rhetoric, Translated by W.Rhys Roberts and test the central proposition of the impact of PDB on consumers’ pref- Ingram Bywater, New York: Modern Library. (Original work erence for affective versus cognitive information. published 1354). The pilot study analyzes print ads from US (PDB =40) and Ja- Botti, Simona and Ann L. McGill (2011), “The Locus of Choice: pan (PDB =54), providing ecological support for the hypothesized Personal Causality and Satisfaction with Hedonic and PDB impact. Utilitarian Decisions,” Journal of Consumer Research, 37 Study 1a and 1b demonstrate the main proposition by manipu- (April), 1065-078. lating both PDB (Zhang et al. 2010, study 1c) and information type (i.e. affective vs. cognitive) across different product categories, thus supporting the internal validity of the main proposition.

Deci, Edward L., Richard Koestner, and Richard M. Ryan (1999), Lalwani, Ashok K. and Lura Forcum (2016), “Does a Dollar Get A meta-analytic review of experiments examining the effects You a Dollar’s Worth of Merchandise? The Impact of Power of extrinsic rewards on intrinsic motivation,” Psychological Distance Belief on Price-Quality Judgments,” Journal of bulletin, 125 (6), 627-68. Consumer Research, 43 (May), 317-333. Dhar, Ravi and Klaus Wertenbroch (2000), “Consumer Choice Maimaran, M. and Ayelet Fishbach (2014). If it’s useful and you between Hedonic and Utilitarian Goods,” Journal of know it, do you eat? Preschoolers refrain from instrumental Marketing Research, 37 (February), 60-71. food. Journal of Consumer Research, 41(3), 642-655. Fera, Rae Ann (2014), “The Rise of Sadvertising: Why Brands Ryan, Richard M. and Edward L. Deci (2000), “Self-Determination are Determined to Make You Cry,” http://www.fastcocreate. Theory and the Facilitation of Intrinsic Motivation, Social com/3029767/the-rise-of-sadvertising-why-brands-are- Development, and Well-Being,” American Psychologist, 55, determined-to-make-you-cry#! 68-78. Fishbach, Ayelet and Jinhee Choi (2012), “When thinking about See, Ya Hui Michelle, Richard E. Petty and Leandre R. Fabrigar goals undermines goal pursuit,” Organizational behavior and (2008), “Affective and Cognitive Meta-Base of Attitudes: Human Decision Processes, 118 (2), 99-107. Unique Effects on Information Interest and Persuasion,” Friesen, Justin P., Aaron C. Kay, Richard P. Eibach, and Adam D. Journal of Personality and Social Psychology, 94 (6), 938-55. Galinsky (2014), “Seeking Structure in Social Organization: Winterich, Karen Page and Yinlong Zhang (2014), “Accepting Compensatory Control and the Psychological Advantages of Inequality Deters Responsibility: How Power Distance Hierarchy,” Journal of Personality and Social Psychology, 106 Decreases Charitable Behavior,” Journal of Consumer (4), 590-609. Research, 24 (August), 274-93. Gao, Huachao, Karen Page Winterich, and Yinlong Zhang (2016), Zhang, Yinlong, Karen Page Winterich, and Vikas Mittal (2010), “All that Glitters Is Not Gold: How Others’ Status Influences “Power-Distance Belief and Impulsive Buying,” Journal of the Effect of Power Distance Belief on Status Consumption,” Marketing Research, 47 (October), 945-54. Journal of Consumer Research, 43 (April), 265-81. Heider, Fritz (1958), The Psychology of Interpersonal Relations, New York: Wiley. Johar, Gita Venkataramani, Durairaj Maheswaran, Larura A. Peracchio (2006), “MAPping the Frontiers: Theoretical Advances in Consumer Research on Memory, Affect, and Persuasion,” Journal of Consumer Research, 33 (June), 139- 49. Acknowledging versus Ignoring the Identity-Relevance of Rivalry: Why Endorsing Dual Identities Decreases Spectator Aggression and Downplaying Makes Things Worse Sebastian Uhrich, German Sport University Cologne, Germany Johannes Berendt, German Sport University Cologne, Germany

EXTENDED ABSTRACT behavioral tendencies (“I feel the desire to hurt / inflict pain on fans Spectator sports are an emotional intergroup setting in which of [rival]”, Mackie, Devos, and Smith 2000) and collapsed these animosity tends to become overheated and crowds quickly shift from measures into an index of aggressiveness. Reactance was added as a peaceful to violence. Incidents of spectator aggression mainly occur mediator in studies 2 and 3. Team identification and trait aggression around clashes of rival teams, a problem that is prevalent across dif- serve as controls. ferent sports and countries. While professional sport teams generally The results show that sport spectator aggressiveness in rival promote rivalries to generate interest and boost attendance, they are competitions can be reduced by dual identity statements compared afraid to add further fuel to the fire as hostilities tend to escalate to both downplay statements and remaining silent. The superiority of quickly between rival spectator groups. Countermeasures include dual identity statements over the no-statement control condition indi- public statements that remind spectators of appropriate and inap- cates that the use of public statements can help sports teams to reduce propriate behavior at sports venues. A widespread approach is the spectator aggressiveness surrounding rivalry games. The managerial attempt to play down the importance of rivalry games, with officials practice to deemphasize the importance of rivalry games produced stating that the game is “no war” or “only for three points.” Although even higher levels of spectator aggressiveness than the no-statement the attempt to downplay the importance of the game appears intui- condition. This finding suggests that the widespread use of downplay tively useful, the effects of such statements on spectator aggression statements is not only ineffective but may even be counterproductive. are unclear. Drawing on the intergroup conflict literature, this re- We contribute to the literature by unravelling the theoretical search identifies so-called dual identity statements and argues that mechanisms underlying the effects of the different types of- state such statements are more effective in reducing aggression between ments. Consistent with our theorizing, the superiority of dual identity rival spectator groups than the managerial practice to downplay. The statements can be explained by their ability to strengthen a superordi- present research also examines the mechanisms that underlie the ef- nate identity that spectators share with rival spectators. Theoretically, fects of dual identity statements and assesses the source of the state- rival spectators move from outgroup to ingroup to some extent and, ment as a potential boundary condition. therefore, become less of a target of aggressiveness. A strength of the dual identity approach is that it refers directly Studies 2 and 3 confirm that increased levels of psychological to consumers’ identity, which is the basis of the conflict. A public dual reactance are partially responsible for the counter-intuitive find- identity statement must include two important aspects to decrease the ing that downplay statements increase aggressiveness compared to hostility toward the rival spectator group: 1) it acknowledges the ri- saying nothing. Spectators’ get upset when teams deemphasize the val spectators’ subgroup identities, including their respective distinct importance of rivalry because the conflict with the rival spectators features (e.g., team colors, rituals) and 2) it reframes group bound- marks a crucial part of their identity. Disentangling these causal aries by defining and promoting a common superordinate identity, mechanisms adds to our understanding of how team sport consum- which is accepted and valued by both spectator groups. ers respond to public statements from professional sport teams and The dual identity approach explicitly promotes a superordinate spells a basis to design effective statements. identity that includes the rival spectators. When spectators are en- Interestingly, the source of the statement (ingroup, outgroup or couraged to reconsider previous group boundaries, the context in both groups together) did not have an influence in our studies. Pre- which social comparisons take place will contain a sense of shared vious research provides a potential explanation for this unexpected existence and common identity (e.g., similarity in geography, tradi- finding. Gomez et al. (2008) show that the endorsement of a super- tion). ordinate identity from the outgroup is acceptable if people know that Holding both the subgroup identity and the superordinate identi- fellow ingroup members also accept the superordinate connection. ty reduces intergroup bias by creating a sense of commonality (Horn- Our study participants may have assumed that fellow spectators ac- sey and Hogg 2000). This should reduce the initial ingroup-favoring cepted the superordinate identity and, therefore, the dual identity bias and makes the opposing group less of a target of aggressiveness: statement was perceived favorably even when it came from the rival We also expect the strength of the superordinate identity to me- team’s players. diate the effects of the type of message on spectator aggressiveness. A caveat of our field experimental approach is a higher degree Three famous rivalries in the German soccer league Bundesliga of noise, limited control over the procedures and a higher number of served as the empirical setting to test our hypotheses in experimental inattentive participants compared to laboratory settings. field studies. Study 1 used a pen-and-paper survey among supporters REFERENCES of Dortmund (N = 419; MAge = 27.32 (±11.23), 66% male), while studies 2 and 3 used online surveys among supporters of Brunswick Berendt, Johannes and Sebastian Uhrich (2016), “Enemies with Benefits: The Dual Role of Rivalry in Shaping Sports Fans’ (N = 949; MAge = 42 (±14.03), 89.5% male) and Nuremberg (N = Identity,” European Sport Management Quarterly, 16 (5), 625; MAge = 33.30 (±13.06), 89% male), respectively. In the dual identity condition, the statement in the press article promoted a su- 613-634. perordinate identity by alluding to a shared love of the rival spectator Brehm, Jack W. (1966), “A Theory of Psychological Reactance,” groups for their region and tradition. In the downplay condition, the New York, NY: Academic Press. statement deemphasized the importance of the game. Converse, Benjamin A. and David A. Reinhard (2016), “On To capture the dependent variable, we measured both aggres- Rivalry and Goal Pursuit: Shared Competitive History, Legacy sive affect (“When thinking about [rival] supporters, I feel hate/ Concerns, and Strategy Selection,” Journal of Personality and anger / disgust”, Kteily, Hodson and Bruneau 2016) and aggressive Social Psychology, 110 (2), 191-213.

Fiol, Marlene C., Michael G. Pratt, and Edward J. O’Connor Mackie, Diane M., Thierry Devos, and Eliot R. Smith (2000), (2009), “Managing Intractable Identity Conflicts,”Academy of “Intergroup Emotions: Explaining Offensive Action Management Review, 34 (1), 32-55. Tendencies in an Intergroup Context,” Journal of Personality Fiske, Susan T. (2002), “What We Know Now About Bias and and Social Psychology, 79 (4), 602. Intergroup Conflict, the Problem of the Century,” Current Northrup, Terrell A. (1989), “The Dynamic of Identity in Personal Directions in Psychological Science, 11 (4), 123-128. and Social Conflict,” inIntractable Conflicts and Their Gneezy, Ayelet (2017), “Field Experimentation in Marketing Transformation, ed. Terrell A. Northrup., Stuart J. Thorson, Research,” Journal of Marketing Research, 54 (1), 140-143. and Louis Kriesberg, Syracuse, NY, Syracuse University Press, Gomez, Angel, John F. Dovidio, Carmen Huici, Samuel L. 55-82. Gaertner, and Isabel Cuadrado (2008), “The Other Side of Oppenheimer, Daniel M., Tom Meyvis and Nicolas Davidenko We: When Outgroup Members Express Common Identity,” (2009), “Instructional Manipulation Checks: Detecting Personality and Social Psychology Bulletin, 34 (12), 1613- Satisficing to Increase Statistical Power,” Journal of 1626. Experimental Social Psychology, 45 (4), 867-872. Hornsey, Matthew. J and Michael A. Hogg (2000), “Assimilation Simon, Bernd, Alex Mommert, and Daniela Renger (2015), and Diversity: An Integrative Model of Subgroup “Reaching Across Group Boundaries: Respect from Outgroup Relations,” Personality and Social Psychology Review, 4 (2), Members Facilitates Recategorization as a Common Group,” 143-156. British Journal of Social Psychology, 54 (4), 616-628. Kteily, Nour, Gordon Hodson, and Emile Bruneau (2016), “They Simon, Bernd, Frank Reichert and Olga Grabow (2013), “When See Us as Less Than Human: Metadehumanization Predicts Dual Identity Becomes a Liability Identity and Political Intergroup Conflict via Reciprocal Dehumanization,”Journal Radicalism Among Migrants,” Psychological Science, 24 (3), of Personality and Social Psychology, 110 (3), 343. 251-257. Optionally Green: The Role of Green Attribute Optionality in Influencing Performance Evaluations Bryan Usrey, University of East Anglia, UK Dayananda Palihawadana, Leeds University, UK Charalampos Saridakis, Leeds University, UK Aristeidis Theotokis, Leeds University, UK

EXTENDED ABSTRACT was optional respondents rated the product as higher in performance As firms seek to develop green products, they may find that con- ability than if it was non-optional (F (1, 73) = 5.796, p < .05). sumers are unwilling to alter their consumption behavior due to a Study 1b introduced choice architecture and the two default op- perceived performance reduction. However, past literature has often tionality policies. One hundred and twenty-five respondents were reviewed green products homogenously, whereby a product is either cruited to test a one-factor (green attribute optionality: non-optional green or non-green. In contrast, we state that a product can feature a vs. opt-in policy vs. opt-out policy) between-subjects experiment. green attribute, but that it is not required for the base product to func- All respondents were given a press release for a washing machine, tion (termed green attribute optionality), potentially overcoming the like Study 1a. In the opt-in condition, participants read that the user negative green product performance stigma. “can activate the EcoX technology” compared to the opt-out condi- Building on the innovation locus literature that recognizes the tion that read that users “can deactivate the EcoX technology”. There distinction between core and peripheral innovations (e.g. Gatignon, was no optionality information in the non-optional condition. Next, et al. 2002), we extend the concept of peripheral to include optional- we measured performance evaluations, green evaluations and envi- ity. When optional the attribute may enhance a product’s green ben- ronmental consciousness. An ANCOVA, including environmental efits, but is not required for it to function. Thus, we posit that perfor- consciousness as a control variable, revealed a significant main effect mance risk is not transferred from the attribute to the base product, (F (1, 121) = 4.364, p <.05), whereby opt-in was significantly higher enhancing performance evaluations. than both the opt-out and the non-optional. Moreover, there was no In addition, we state that optionality includes two default poli- significant difference between the non-optional and the opt-out con- cies. Based on choice architecture and default literature, a product ditions. Finally, no significant effect was found on green evaluations may possess a green attribute that is default, but may be deactivated (p = .128). (opt-out). Alternatively, the green attribute may be deactivated, al- Study 2 introduces the moderating variable of cognitive style. lowing the user to activate it when needed (opt-in). As defaults act as We recruited 245 respondents to examine a 3 x 2 between-subjects a carrier of meaning (Brown and Krishna, 2004), we posit that when experiment. The manipulations for optionality were the same as in the green attribute is the default option (opt-out), performance evalu- Study 1b, while cognitive style was manipulated by asking respon- ations will be reduced. dents to write about a meaningful event that took place in their lives Additionally, we explore the moderating role of cognitive style alone (analytical) or with friends and/or family (holistic). A signifi- and the mediating variable of green product typicality. In the former, cant main effect was found (F (1, 238) = 9.887, p < .05), whereby we posit that holistic thinkers will view product attributes as inter- opt-in enhanced performance evaluations. In addition, a significant connected, focusing on the degree to which differing attributes are interaction was found (F = (1, 238) 4.363, p < .05), where the analyti- connected and form an entity. In contrast, analytical thinkers may cal condition negated the effect of choice architecture. Finally, there be able to detach the attribute from the product, viewing each at- was no main effect for green evaluations (p = .764). tribute as separate (Nisbett, et al., 2001). Based on this, we posit that Finally, in Study 3, we test a 2 x 2 between subjects’ experiment the impact of green attribute optionality on performance evaluations with 149 respondents to examine the mediating variable of green will be negated when an analytical mindset is activated. In the latter, product typicality, which was measured along with performance and we predict that an opt-out strategy, whereby the green attribute is green evaluations and environmental consciousness. In this study, default, will lead to an increase in the extent to which the product is the non-optional condition was removed. Cognitive style was ma- viewed as typical of similar green products, thus degrading perfor- nipulated as in Study 2, while optionality was manipulated via an ad- mance evaluations. vertainment for a kettle, in which the green was described similarity Three experiments were conducted to examine the effect of to Study 2. A two-way ANCOVA revealed a significant interaction green attribute optionality on performance evaluations. In Study 1a, between green attribute optionality policy and cognitive style (F (1, we explore both the optional and non-optional conditions. In Study 144) = 4.690, p = <. 05). Furthermore, a significant direct effect was 1b, we introduce the default policies, investigating opt-in, opt-out found (F (1, 144) = 5.014, p <.05). Finally using Process Model 7 and non-optional strategies. In Study 2, we examine the moderating (Preacher and Hayes, 2008), we found a significant moderated me- variable of cognitive style. Finally, in Study 3, we present the mediat- diation with a 95% confidence interval excluding zero (CI = -.3895 ing mechanism of green product typicality. to -.0029). In Study 1a, we performed a one-factor (green attribute option- Taken together, we show that optionality enhances performance ality: non-optional vs. optional) between subjects’ experiment with evaluations. Specifically, the optionality policy of opt-in. In addi- 75 respondents. Green attribute optionality was manipulated by pre- tion, this perceived performance enhancement does not come at the senting respondents with an advertisement for a washing machine expense of perceived greenness. Additionally, we find support for along with a heading that outlined the product’s eco-friendly mode. the moderating variable of cognitive style and the mediating role of Located at the bottom of the advertisement, a statement indicated green product typicality. Our results provide both theoretical impli- that the eco-mode was user-activated. In the non-optional condition, cations, as well as practical contributions for green product develop- no statement was presented. Finally, we measured performance eval- ment. uations. An ANOVA revealed that when the environmental attribute

Advances in Consumer Research 926 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 927 REFERENCES Nisbett, Richard E, Kaiping Peng, Incheol Choi, and Ara Brown, Christina L, and Aradhna Krishna. (2004). “The Skeptical Norenzayan. (2001). “Culture and Systems of Thought: Shopper: A Metacognitive Account for the Effects of Default Holistic versus Analytic Cognition.” Psychological Review, Options on Choice.” Journal of Consumer Research, 31 (3), 108 (2), 291-310. 529-539. Gatignon, Hubert, Michael L. Tushman, Wendy Smith, and Philip Anderson. (2002). “A Structural Approach to Assessing Innovation: Construct Development of Innovation Locus, Type, and Characteristics.” Management Science, 48 (9), 1103-1122. The Role of Women, Sexualization and Objectification in LGBTQ Advertising Patrick van Esch, Moravian College, USA Ally Geisler, Moravian College, USA Lizabeth Kleintop, Moravian College, USA Gavin Northey, University of Auckland, New Zealand Jonas Heller, University of New South Wales, Australia

EXTENDED ABSTRACT the underlying LGBTQ message from the non-LGBQ population In advertising, the sexualization of women is taking place in (Peňaloza, 1996). the LGBTQ target market segment. Whilst LGBTQ advertisements continue to target males using humor as the dominant appeal; for Hypothesis 2: For female characters in LGBTQ TV advertising women, the most dominant appeal is sexual. When the main charac- campaigns, the most dominant appeal is sexual ter in the advertisement is portrayed in terms of LGBTQ connotation, The study examined the characteristics of LGBTQ TV adver- a positive and negative connotation is fairly evenly distributed for tisements (n = 300) that ranged from the years 1952 to 2016. It ex- male characters, however; for female characters, the positive LG- amined the portrayal of the main character and voiceover (gender, BTQ connotation is dominant. If the main character is female, she is sexuality, antagonist, protagonist), dominant appeal (emotional, fear, more likely to be portrayed as a protagonist. humor, rational, sexual) and LGBTQ connotations (positive, nega- Marketers and organizations alike, balance a fine line between tive). appealing to both LGBTQ and mainstream clients without ostraciz- When comparing the Western world (US, EU, Oceania) and less ing one or the other, specifically their mainstream clients, as they developed countries (Asia, Africa, South America), LGBTQ adver- are a larger portion of the population (Flounders & Mabry-Flynn tisements continue to target males (χ2 = 4.74, df = 1, p < .05). The (2016). Heterosexuals desire lesbian imagery over gay male imag- broader focus remains on the Western world (135 ads, 46.7% male; ery, therefore, advertisers targeting the gay market will use lesbian 50 ads, 17.3% female), whilst similar in less developed countries (63 imagery because not only does it reach homosexual consumers, but ads, 21.8% male; 41 ads, 14.2% female), the gap between genders is also reduces the chance of heterosexuals identifying the advertise- much smaller. ment as LGBTQ (Hootens et. al., 2009). The heterosexual audience If the main character is male, then the voiceover is traditionally has a positive attitude towards lesbian ads, and these results are com- male as well (87%), hence it is congruent (χ2 = 18.59, df = 1, p < pounded when the level of intimacy displayed in the advertisements .001). For females, the voiceover gender is more evenly distributed are amplified (Oakenfull & Greenlee, 2005). It is significant to note (54.3% male, 45.7% female). However only very few cases occur that lesbians are often depicted as highly sexualized in mainstream where advertisements have a female voiceover (33 ads, 19.9%). In media (Flounders & Mabry-Flynn, 2016). terms of how the main character of the advertisement is portrayed This exploratory research examines the the portrayal of women with regards to LGBTQ connotation, a positive and negative con- when either the main character and/or voiceover (gender, sexuality, notation is fairly evenly distributed for male characters, however, for antagonist, protagonist) in LGBTQ TV advertisements. In the first female characters (36 ads, 62.1%), the positive LGBTQ connotation instance, the aim was to determine the dominant appeals (emotional, is dominant (χ2 = 12.92, df = 2, p < .01). Lastly, if the main character fear, humor, rational, sexual) and LGBTQ connotations (positive, is female (46 ads, 79.3%), she is more likely to be portrayed as a negative) that are associated with women in advertisements to the protagonist (χ2 = 6.32, df = 1, p < .05). LGBTQ community. The sexualization of women also takes place in the LGBTQ The diversity of the LGBTQ community has meant there isn’t community (χ2 = 12.94, df = 4, p < .01). For male characters, the one, single ‘LGBTQ’ identity. Oakenfull (2007) found clear distinc- most dominant appeal is humor (80 ads, 36.5%), followed by sexual tions between gay and lesbian consumers and argued that marketers appeals (56 ads, 25.6%). For female characters, the most dominant had to use lesbian imagery and not just gay imagery in advertise- appeal is sexual (24 ads, 41.4%), with all other appeals evenly dis- ments to be effective in the two communities. Um (2014) found het- tributed. If the character is male (175 ads, 63.2%), the target audi- erosexual consumers to respond less favorably to gay themed adver- ence is more likely to be straight rather than identify as LGBTQ (χ2 tising, including a differential in heterosexual males more negatively = 7.57, df = 1, p < .01). evaluating gay-themed advertising and the brand advertised than did The dominant appeals in LGBTQ advertisements are humor, heterosexual females. Oakenfull & Greenlee (2004) cautioned mar- eroticism, and leisure. Reichert & Ramirez (2000) discovered four keters to consider both the gender of their target gay or lesbian audi- predominant traits of sexual advertisements including: physical ap- ence and the gender of the heterosexual audience likely to view the pearance of models, behavior or movement, intimacy between mod- advertising when creating their advertisements. els, and contextual features. Gill (2009) argues that the proliferation Hypothesis 1: Women are featured less prominently in LGBTQ of sexually explicit imagery within both contemporary and LGBTQ TV advertising campaigns advertising indicates that sexualization is neither a homogenous or singular process. Individuals within the LGBTQ community are Using sex as a method of selling supports the influence of si- sexualized in specific ways and in a visual culture, lesbian, bisexual, lencing a woman’s’ appeal by exhibiting women in an objectified and queer women are often easily recognizable ‘figures’ in LGBTQ manner for male pleasure and utilization. This common portrayal advertising, yet sexualization doesn’t necessarily maneuver outside accentuates physical attractiveness, prominent displays of the body, of class, gender and/or radicalization due to the fact that it remains and sexual mannerisms of women. Then, marketers used more ex- overwhelmingly ageist and heteronormative (Evans, Riley & Shan- plicit messages, though this was imagery that was unique and recog- kar, 2010). In this respect, findings from the current research are pos- nizable to the LGBTQ community, these methods were used to mask sibly the first evidence that women are underrepresented in LGBTQ advertising and female characters are often portrayed with a positive Advances in Consumer Research 928 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 929

LGBTQ connotation. The perspectives of the sexualization of wom- Oakenfull, G.K. & Greenlee, T.B. (2004). The three rules en in LGBTQ advertising highlights two key areas for consideration of crossing over from gay media to mainstream media (1) the portrayal of women compromising the relationship they have advertising: lesbians, lesbians, lesbians. Journal of Business with their own and other women’s bodies, and (2) LGBTQ advertis- Research, 57, 1276-1285. ing sensibility, underwrites the discussion around the sexualization Oakenfull, G.K. & Greenlee, T.B. (2005). Queer eye for a gay guy: of women, including the prospects of empowering female sexuality Using market-specific symbols in advertising to attract gay rather than demonizing it (Randazzo, Farmer & Lamb, 2015). consumers without alienating the mainstream. Psychology & Marketing, 22(5), 421–439. REFERENCES Oakenfull, G.K., M.S. McCarthy, and T.B. Greenlee. 2008. Borgerson, J. L., Schroedera, Jonathan E., Blomberg, B., and Targeting a minority without alienating the majority: Thorssénc, E. (2006a). The Gay Family in the Ad: Consumer Advertising to gays and lesbians in mainstream media. Journal Responses to Non-traditional Families in Marketing of Advertising Research, 48, 191–198. Communications. Journal of Marketing Management, 22, Peňaloza, L. (1996). We’re here, we’re queer, and we’re going 955-978. shopping! A critical perspective on the accommodation Borgerson, J.L., Isla, B., Schoeder, J.E., and Thorssen, E. (2006b). of gays and lesbians in the U.S. marketplace. Journal of Representation of gay families in advertising: Consumer Homosexuality, 31(1-2), 9-41. responses to an emergent target group. In K. M. Ekstrom & H. Pounders, K., & Mabry-Flynn, A. (2016). Consumer Response to Brembeck (Eds.), European Advances in Consumer Research, Gay and Lesbian Imagery. Journal of Advertising Research, 7, 143-152, Goteborg, Sweden: Association for Consumer 56(4), 426-440. Research. Randazzo, R., Farmer, K., & Lamb, S. (2015). Queer Women’s Cunningham, George B. and Melton, E. Nicole. (2014). Signals and Perspectives on Sexualization of Women in Media. Journal of Cues: LGBTQ Inclusive Advertising and Consumer Attraction. Bisexuality, 15(1), 99-129. Sport Marketing Quarterly, 23(1), 37-46. Reichert, T., & Ramirez, A. (2000). Defining sexually oriented Evans, A., Riley, S., & Shankar, A. (2010). Technologies of appeals in advertising: A grounded theory investigation. sexiness: Theorizing women’s engagement in the sexualization Advances in Consumer Research, 27. of culture. Feminism & Psychology, 20(1), 114-131. Rivendell Media, Inc. 2007. 2006 Press Report. Retrieved 7 August Gentzkow, M. & Shapiro, J.M. (2006). Media bias and reputation. 2008 from http://rivendellmedia.com/gay_pressreports/ Journal of Political Economy, 114(2), 280-316. GayPressReport_2006.pdf. Gill, R. (2009). Beyond the sexualization of culture thesis: An Roth, Tom & Paisley, David. (2017). LGBTQ Marketing: What intersectional analysis of sixpacks’, midriffs’ and hot lesbians’ makes us different? What makes us the same?Out & Equal in advertising. Sexualities, 12(2), 137-160. Workplace Advocates 2017 Virtual Summit Series. Webinar Ginder, W. & Byun, S. (2015). Past, present, and future of gay and retrieved from https://cc.readytalk.com/cc/playback/Playback. lesbian consumer research: Critical review of the quest for the do?id=5dp722 Queer dollar. Psychology & Marketing, 32(8), 821–841. Tuten, T. L. (2006). Exploring the Importance of Gay-friendliness Hooten, M. A., Noeva, K., & Hammonds, F. (2009). The effects of and its Socialization Influences.Journal of Marketing homosexual imagery in advertisements on brand perception Communications, 12(2), 79–94. and purchase intention. Social Behavior and Personality: An Um, Nam-Hyun. (2014). Does gay-themed advertising haunt International Journal, 37(9), 1231-1238. your brand? The impact of gay-themed advertising on young Lambiase, J., & Reichert, T. (2003). Promises, promises: Exploring heterosexual consumers. International Journal of Advertising, erotic rhetoric in sexually oriented advertising. Persuasive 33(4), 811-832. imagery: A consumer perspective, 247-266. Um, Nam-Hyun. (2016). Consumers’ Responses to Implicit and Oakenfull, G.K. (2007). Effects of gay identity, gender and Explicit Gay Themed Advertising in Gay vs. Mainstream explicitness of advertising imagery on gay responses to Media, Journal of Promotion Management, 22(3), 461-477. advertising. Journal of Homosexuality, 53(4), 49-69. Personal Control and Perceived Product Efficacy Noah VanBergen, University of Cincinnati, USA Caglar Irmak, University of Miami, USA

EXTENDED ABSTRACT control increased perceived efficacy in the neutral affirmation condi- A growing stream of research examines how individuals cope tion (p<.05) but not the self-affirmation conditionF ( <1). with the common experience of reduced personal control (e.g., Chen, Study 2 built on these findings by manipulating the personal Lee, and Yap 2016; Cutright 2012; Cutright and Samper 2014). Com- relevance of the product under consideration, such that half the par- pensatory control theory (CCT; Kay and Eibach 2013) proposes that ticipants believed the product under consideration would be used for behavior following a loss of control is driven by to two basic moti- an upcoming task, whereas the other half merely reported efficacy vations: bolstering a sense of agency, and affirming structured and perceptions without expecting to consume the product or perform predictable relationships in the world. In turn, four strategies serve the task. Thus, 209 undergraduates were randomly assigned to one of these two motivations (Landau, Kay, and Whitson 2015). The pres- four conditions in a 2 (personal control: low vs. high) x 2 (expecta- ent research builds on the two strategies for bolstering agency, which tion of consumption and performance: absent vs. present) between- are (1) affirming beliefs about personal agency or ability, and (2) subjects design. Participants in the expectations-present condition affiliating with external entities that can act on the self’s behalf. Spe- were led to believe that they would consume an energy drink (“Neuro cifically, we build on prior work showing that control-deprived con- SONIC”) before completing a performance task, while those in the sumers evaluate products more favorably when those products have expectations-absent condition were simply informed that we were in- characteristics fitting the motivation to affirm personal agency (i.e., terested in their perception of the energy drink. We measured percep- “strategy 1”). For example, products requiring effort are evaluated tions of product efficacy, as well as participants’ motivation to derive more favorably because using them is empowering (Cutright and efficacy from the product in order to accomplish desired outcomes. Samper 2014), and utilitarian products are evaluated more favorably We hypothesized that low (vs. high) personal control would increase because they are associated with problem-solving (Chen et al. 2014). perceived efficacy in the expectations-present condition only, and Extant research has not examined the possibility that products that motivation to derive product efficacy would mediate this effect. can serve as external sources of control (i.e., “strategy 2”); research Results confirmed our predictions. We found significant two- in this strategy focuses on cultural sources of control (e.g., religion way interactions on both efficacy perceptions (p<.05) and motivation and governments; Kay and Eibach 2013). However, because “ef- (p<.05), such that participants in the expectations-present condition ficacious products” (e.g., performance-enhancers, detergents, and perceived significantly greater efficacy, and were also significant- pharmaceuticals) can be seen as external entities that act on a user’s ly more motivated to derive efficacy, under low vs. high personal behalf in order to bring about a desired outcome (e.g., successful control (ps<.05). Mediation analysis (Hayes 2013; Model 8) dem- performance, a clean home, and relief from ailments, respectively), onstrated that motivation mediated the effect of personal control on it may be possible for control-deprived consumers to rely on such efficacy perceptions (95% CIs for both the two-way interaction and products as external sources of control. simple effects did not cross zero, and the direct effect of the two-way In line with CCT, we thus propose that consumers experienc- interaction on efficacy became nonsignificant when accounting for ing low control seek external aids that can help them achieve their motivation). goals and reaffirm their ability to bring about desired outcomes. As Our research contributes to the personal control and product efficacious products can help users reach desired outcomes, we hy- efficacy literatures. First, extant marketing research on products as pothesize that low (vs. high) personal control increases perceived sources of personal control has studied only evaluations of prod- product efficacy. Moreover, as this is driven by control-deprived ucts, showing that products possessing characteristics associated consumers’ motivation to use the product in order to reach a desired with regaining agency are evaluated more favorably (Chen et al. outcome, this effect combines “strategy 2” and “strategy 1” (Landau 2016; Cutright and Samper 2014). We build on this work by show- et al. 2015) in the sense that reduced control leads consumers to give ing that perceptions, not only evaluations, of products are impacted more credit to the external source (by perceiving greater efficacy) by personal control. Moreover, this effect is due to a combination because this helps consumers accomplish desired outcomes through of the strategies used to regain control: Control-deprived consum- their own actions. This implies that the effect of personal control on ers perceive greater product efficacy because they are motivated to perceived product efficacy should be moderated by the personal rel- utilize external sources of control (“strategy 2”) in order to accom- evance of a product for achieving desired outcomes. plish personally-relevant desired outcomes (“strategy 1”; Landau et We tested these predictions in two studies and find support for al. 2015). In contrast, previous research has only shown that product our proposed process. Study 1 provides initial evidence through a 2 evaluations are impacted by personal control due to “strategy 1.” Fi- (personal control: low vs. high) x 2 (self-affirmation: neutral vs. af- nally, we also contribute to the product efficacy literature by showing firmation) between-subjects design in which 98 Mturk respondents that personal control—a factor relating to consumers, rather than the participated. As self-affirmation reduces the need to compensate for product—impacts perceptions of product efficacy. low personal control (Whitson and Galinsky 2008), we hypothesized that participants feeling low (vs. high) personal control would per- REFERENCES ceive a product as more effective, but this effect would be attenuated Chen, Charlene Y., Leonard Lee, and Andy J. Yap (2016), by an opportunity to self-affirm. To test this, participants were led to “Control Deprivation Motivates Acquisition of Utilitarian believe that they would utilize a performance-enhancing treatment Products,” Journal of Consumer Research, ucw068. (listening to music by Mozart) to assist their performance on a task, Cutright, Keisha M. (2012), “The Beauty of Boundaries: When and we measured participants’ expectations for the effectiveness of and Why We Seek Structure in Consumption,” Journal of the music at improving performance. As expected, we found a sig- Consumer Research, 38 (5), 775–90. nificant 2-way interaction (p<.05) such that low (vs. high) personal

Cutright, Keisha M. and Adriana Samper (2014), “Doing It the Landau, Mark J., Aaron C. Kay, and Jennifer A. Whitson (2015), Hard Way: How Low Control Drives Preferences for High- “Compensatory Control and the Appeal of a Structured Effort Products and Services,”Journal of Consumer Research, World,” Psychological Bulletin, 14 (3), 694-722. 41 (3), 730–45. Whitson, Jennifer A. and Adam D. Galinsky (2008), “Lacking Hayes, Andrew F. (2013), An Introduction to Mediation, Control Increases Illusory Pattern Perception,” Science, 322 Moderation, and Conditional Process Analysis: A Regression- (5898), 115–17. Based Approach, New York: Guilford. Kay, Aaron C. and Richard P. Eibach (2013), “Compensatory Control and Its Implications for Ideological Extremism,” Journal of Social Issues, 69 (3), 564–85. How Process Explanations Impact Assessments of Predictions of Uncertain Events Daniel Villanova, Virginia Tech, USA Elise Chandon Ince, University of South Carolina, USA Rajesh Bagchi, Virginia Tech, USA

EXTENDED ABSTRACT tions, separate ANOVAs with trust and accuracy as dependent vari- When making predictions, experts can choose how much ex- ables revealed a significant effect of level of detail on accuracy and planation to provide—either very little or a lot—about the process a marginal effect on trust. The patterns of the effects were similar. used to make inferences. While literature on consumer research sug- For both accuracy and trust, compared to providing no explanation, gests that providing more information generally improves evalua- providing a brief explanation reduced accuracy and trust, whereas tions (Calder, Insko, and Yandell 1974; Chaiken 1980; Stiff 1986), providing a detailed one did not. we show that consumers report preferences for more explanation in Thus, providing a brief explanation resulted in reduced accu- choices among predictions, but, in the context of single probability racy and trust. Study 3 replicates these effects in a stock price con- prediction evaluations, we demonstrate that providing process in- text, demonstrates the effect on two more downstream variables, and formation can lower evaluations. This occurs because explanation examines the mediating role of perceived depth of analysis. sometimes leads to the inference that the prediction-making process In study 3, participants (N=259) from Amazon’s MTurk paris less complicated, and, therefore, the expert’s analysis was not as ticipated in this study for monetary compensation. The analyst had in-depth, affecting inferences about both the prediction (accuracy) predicted the chance that a particular company’s stock price would and the expert (trustworthiness). Thus, consumers may believe a pre- increase by $10 per share in the next year. diction to be less accurate and trust the expert less when she offers Two of the choice conditions were the same as previous studies. more explanation. We also added a third choice condition in which Analyst A offered We draw from the area of inference-making to develop our a brief explanation, and Analyst B offered a detailed explanation. In theory. This research stream suggests that when little information is the evaluation conditions, participants evaluated a prediction with present (e.g., other than the overall prediction), consumers use prior either no, a brief, or a detailed explanation. knowledge to make inferences. Because an expert provided this es- Participants in the choice conditions then chose an analyst’s timate, they infer that the process of arriving at the overall predic- prediction they would use. Then they stated which analyst they tion was complicated. However, when an explanation is presented, would be more likely to recommend and responded to the same trust consumers can generate new information from this content. Hence, and accuracy measures as in study 2, plus a new depth of analysis when a brief explanation is presented, it is likely to lack details and item. In the evaluation conditions, except for choosing the analyst, look superficial, leading to inferences that the expert did not do an participants rated the same questions, adapted accordingly, and also in-depth analysis, which should be less likely to occur with detailed indicated how likely they would be to purchase the stock. explanations. The results mirrored those of the previous studies- there was In study 1, participants (N=121) from Amazon’s Mechanical a preference for explanation in choice, but brief explanations were Turk participated in this study for monetary compensation. Partici- punished in evaluations. Depth of analysis mediated the effect of pants were asked to choose one of two analysts’ predictions to share brief explanation on the four evaluation variables. with others. Across three studies, using job and stock prediction contexts, They were told that the analysts had made predictions about we demonstrate the role that explanation plays in affecting choices job prospects in various fields, one of which related to finding ad- and evaluations. We provide insights into the role of information in ministrative jobs in the hoteling and tourism (HT) industry. Partici- information products, with the counterintuitive finding that experts pants chose between an expert who did not provide any explanation may be better off leaving their predictions unexplained. (Analyst A) and one who did (Analyst B). In the brief explanation condition, Analyst B said he looked at two factors, the chances of REFERENCES finding an administrative job and finding a job in the HT industry, Calder, Bobby J., Chester A. Insko, and Ben Yandell (1974), “The and combined them. In the detailed explanation condition, Analyst B Relation of Cognitive and Memorial Processes to Persuasion in described four steps used to arrive at the same two factors and in the a Simulated Jury Trial,” Journal of Applied Social Psychology, fifth step combined them. 4 (1), 62-93. Participants overwhelmingly (87%) chose the analyst that pro- Chaiken, Shelly (1980), “Heuristic versus Systematic Information vided an explanation, regardless of the level of detail provided by the Processing and the Use of Source versus Message Cues in explanation. Study 2 investigates what happens in a single evaluation Persuasion,” Journal of Personality and Social Psychology, 39 context. (5), 752-766. In study 2, participants (N=243) from Amazon’s MTurk par- Garfield, Joan and Andrew Ahlgren. (1988), “Difficulties in ticipated in this study for monetary compensation. The scenario was Learning Basic Concepts in Probability and Statistics: similar to that of study 1. In the choice conditions, participants chose Implications for Research,” Journal for Research in between experts. In the evaluation conditions, participants evaluated Mathematics Education, 19 (1), 44-63. a prediction with either no, a brief, or a detailed explanation. Stiff, James B. (1986), “Cognitive Processing of Persuasive Participants in the choice conditions then chose an ana- Message Cues: A Meta-analytic Review of the Effects of lyst’s prediction they would use, followed by evaluations of Supporting Information on Attitudes,” Communication their relative trust in the analysts and the accuracy of the predic- Monographs, 53 (1), 75-89. tions. Instead of choosing an analyst, participants in the evaluation conditions completed measures of trust and accuracy. The choice results mirrored those of study 1. In the evaluation condi-

EXTENDED ABSTRACT often unable to understand cultural aspects of marketplaces (Taylor Largely excluded from marketing literature are individuals in and Stanovic 2014). subsistence marketplaces. This term refers to an approach that begins Abject poverty, arising from loss of income and assets, and dis- at the micro-level and is bottom-up, encompassing a range of low- placement, make markets inaccessible and products unaffordable, income consumers and entrepreneurs (Viswanathan 2013). Indeed, severely inhibiting consumption capability (Deaton 1972). Poverty these marketplaces are often characterized by exclusion. But market- is rampant among refugees who have lost their livelihoods in their place exclusion reaches even starker levels in subsistence contexts home countries (Sartorius et al. 2013). Refugees mostly survive on borne out of violence, flight, and abject conditions for survival, such government or international aid with limited alternative means to as in refugee settlements. Understanding such exclusion is especially their survival. The social and psychological conditions refugees face critical as a refugee’s average length of stay in exile is approximately make it extremely difficult for them to start a new livelihood. This 20 years (Betts et al. 2014). is coupled with the lack of economic opportunities made available We aim to understand how refugees are excluded from the mar- to them until their legal status is settled, which can take many years ketplace. These experiences are so drastic as to stretch extant con- (White 2007). ceptualizations to cover extreme marketplace exclusion. The funda- Loss of marketplace literacy due to relocation makes it difficult mental contribution of this paper is therefore to delineate extreme to engage in marketplace interactions (Viswanathan et al. 2009). Loss marketplace exclusion via refugee contexts. We explain such exclu- of such literacy can lead to poor consumption decisions, sometimes sion by integrating current research and providing empirical support making individuals vulnerable to deception at the hands of sellers. for our conceptualization. Our synthesis of extant literature identifies Whereas in most subsistence contexts, consumers cope with the lack elements of extreme marketplace exclusion and their interplay. We of basic literacy by developing marketplace literacy via interpersonal then describe our methods and present findings about the nature of relations (Viswanathan et al. 2009), refugees face a new, unfamil- extreme marketplace exclusion in our research context. We conclude iar and extremely harsh setting, where social relations are strained with a discussion covering theoretical and practical implications. (Werker 2007). The loss of marketplace literacy is accentuated for those with low literacy and income, whose functional literacy in the Elements of Extreme Marketplace Exclusion marketplace domain is socially embedded. For those with higher lit- Literature on exclusion has primarily focused on ‘social’ exclu- eracy, there is still an effect due to situational loss of marketplace lit- sion (Sen 2000), examining it as a vicious cycle with multiple, recip- eracy (Viswanathan et al. 2010). The psychological stresses of leav- rocal, cause-and-effect elements which intensify exclusion (Levitas ing home countries and resettlement may further impair refugees’ 1996). Marketplace exclusion may have varying degrees, with ‘deep cognition, characteristic of low-income consumers living in scarcity exclusion’ characterizing exclusion across multiple dimensions conditions (Shah, Mullainathan, and Shafir 2012) Given the link be- (Levitas et al. 2007). Similar to deep exclusion, extreme marketplace tween marketplace literacy and self-confidence (Viswanathan et al. exclusion involves multiple forces. Unlike deep exclusion, however, 2009), we consider emotions and cognitions. Emotions are drastic in we study exclusion in a subsistence marketplace context intensified refugee contexts, even more so than other subsistence marketplaces by violence, psychological stress, economic strain, and social insta- and a violent past (and sometimes present) combined with deep fear bility. Specific to a refugee context, our research focuses on how of the future fixes individuals in the immediate present, accentuating elements relate to form extreme marketplace exclusion, locating it cognitive elements. vis-a-vis deep exclusion. Importantly, the marketplace does not only involve economic transactions, but melds social, cultural, political Context and Methods and economic realities (Granovetter 1990). Research remains to offer Context. We conducted our study in Nakivale, a refugee settle- a theoretical model to account for these multiple elements of mar- ment in rural Uganda housing over 84, 000 refugees managed by the ketplace exclusion. We now identify and delineate four elements of United Nations High Commissioner of Refugees (UNHCR). Upon this type of marketplace exclusion: geographic isolation, social os- arrival, refugees are given construction materials to build shelter. tracism, abject poverty and loss of marketplace literacy. Residences are allocated based on ethnicity due to potential conflict Studies in rural subsistence marketplaces (e.g., China; Wang among refugees of different nations. Between the time of arrival and and Tian 2013) in refugee settlements (Rose 2001) show geographic time spent building, they sleep under a large tent. The UNHCR dis- isolation as a key contributor to marketplace exclusion. Many settle- tributes a limited supply of food each month, but it is gradually re- ments lack sufficient infrastructure, geographically isolated from duced after six months with the expectation that refugees find work. public facilities including schools, hospitals, and markets (De Mont- Employment is often difficult to obtain, however. Mbarrara, a mid- clos and Kagwanj 2000). Even more drastic, refugees’ geographic sized town approximately 45 minutes away from Nakivale by car, isolation may stem from legal restrictions enforcing segregation offers employment opportunities, but no formal transportation to the from the local community (Taylor 1996). town exists. We also note these refugees vary widely in their educa- Refugees in their host countries often face the most severe form tional achievements, different from traditional subsistence market- of social ostracism, as they bear the stigma of fleeing their home places where low literacy and low income often co-occurs (Hodes country, and are perceived as unwelcome exploiters of the host and Goldberg 2002). Images depicting the severe circumstances of country’s resources (Sartorius et al. 2013). Refugees must endure the marketplace can be found in the Appendix. the turmoil of resettlement, separation from family and culture (Sch- Methods. We employed qualitative techniques including obser- weitzer, Greenslade, and Kagee 2007), and exclusion due to cultural/ vations, depth interviews, and marketplace-learning interventions. language barriers (Taylor and Stanovic 2004). Refugees are therefore Pertaining to field observations, two immersions took place in person (one in June 2015 and again in July 2016), and virtually from January

Advances in Consumer Research 933 Volume 45, ©2017 934 / Extreme Marketplace Exclusion in Subsistence Marketplaces: A Study in a Refugee Settlement in Nakivale, Uganda to May, 2015 via Skype meetings. The UNHCR and non-profit orga- from the home country or of different origin often exclude refugees nizations on the ground facilitated access. In particular, we worked via discrimination. One informant illustrates this: with a group of entrepreneurial refugees mentored and developed into a team by one of our partner organizations. These refugees acted Daniel: … clearly I can speak English and those sellers can as both participants and facilitators. We observed settlement facili- speak English, but when in Mbarara, they know I’m not Ugan- ties, various local marketplaces, the land, and a nearby town. In- dan – and they can change and instead they don’t speak Eng- terviews ranged from 30 to over 90 minutes and tape recorded for lish… just to make a complication. Here, I used a different lan- coding purposes. The first part of each semi-structured interview was guage, I do not negotiate and sometimes they increase price. phenomenological (Thompson, Locander, and Pollio 1989), and the second transitioned into a structured discussion of informants’ mar- Daniel explains how social ostracism manifests in the market- ketplace experiences. Refugee participants and translators received place. The exclusion he experiences cannot be neatly categorized into 3,000 shillings for participation, the equivalent to one U.S. dollar. being indirectly ignored or directly rejected (Lee and Shrum 2012). The shopkeepers clearly acknowledge him (he is not ignored), and Findings they do not explicitly refuse his shopping ability (he is not directly We examine four elements of extreme marketplace exclusion: rejected). Instead, shopkeepers acknowledge him in an attempt to co- abject poverty (economic), geographic isolation (physical), social erce a ‘penalty price’ for not being Ugandan. Consequently, the sell- ostracism (social), and loss of marketplace literacy (cognitive-emo- ers exclude the refugees from fully participating in the marketplace. tional). Importantly, extreme marketplace exclusion impacts both Loss of Marketplace Literacy. Another element leading to consumption and employment. extreme marketplace exclusion is the loss of marketplace literacy, defined as encompassing skills, self-confidence, and awareness of Elements of Extreme Marketplace Exclusion rights and capturing know-why, know-how, and know-what as con- Abject poverty. In terms of economic conditions, refugees face sumers and entrepreneurs (Viswanathan et al. 2009). We note that extreme poverty. Finding food of some quality and clean water is refugees may range in income and literacy levels in their home coun- challenging. Kevin, a middle-aged refugee employed at an UNHCR try, with the refugee status drastically lowering the former while facility, states that the “Sometimes [the] water has color, sometimes maintaining the latter for relatively higher incomes and education it smells, and good water doesn’t have all those elements.” Lacking levels. We focus on cognitive and emotional dimensions. even basic necessities leaves refugees with few spending options. A Concrete thinking in the here and now arising from difficulty young man describes his experience of abject poverty and how he with abstractions is likely to be accentuated with geographic iso- spends in the local marketplace. lation, abject poverty, and social ostracism. Similarly, pictographic thinking and a rudimentary dependence on senses is also particular- David: We have nowhere to borrow from. ized in the immediate. Veronica, a 20-year-old woman, illustrates her Interviewer: Do you buy things in the market? present-state orientation explaining, “I don’t have savings…” and David: … not much. We get soap and cooking oil. when discussing leisure activities, “Never I have the time to think of that.” Mary, a 43-year-old woman, explains how difficult it is to fall Our informant explains how he has no accessible credit institu- asleep. She elaborates, “I don’t get sleep thinking how will I push the tions to borrow funds. Due to his scarce financial resources, he buys next day and make living for my children.” Cognitively, fathoming a only necessities for the immediate present. Unlike traditional sub- future is difficult for these women. Emotionally, Veronica describes sistence contexts, any earned/inherited family assets are completely how she cannot engage in leisure activities. Ray even explains how lost upon migration. he attempts to actively repress his emotions just to function saying Geographic Isolation. Although refugees can travel to towns “Such [bad feelings] you cannot permit. You know we are human nearby, the lack of resources and mobility barriers are often insur- beings, yeah? When something is very bad, you [will] have some mountable. Daniel, a 24-year old man who fled Congo and lived in emotions, but maybe [they] can cause you trauma…” The cognitive the settlement since 2010, explains how this particular element im- predilections of being in the here and now arising from low literacy pacts his work as an artist. and the immediacy of basic needs are further accentuated by severe Daniel:…as an artist, I want to buy [tools] I [need]…. [but] emotional trauma. there’s no place they sell [the art materials]…. once I sent some- Interplay between Elements of Extreme Marketplace one to buy [the materials in Kampala].…[but] he doesn’t know Exclusion the quality that I need…. he may just buy what [sellers] tell him. The economic, geographic, social and cognitive-emotional realms of life described all amplify and reinforce extreme market- Due to long distances and high transportation costs, he cannot place exclusion. In a context of abject poverty, geographic isolation even travel to the preferred marketplace in Kampala (260 kilometers precludes opportunities outside the settlement, narrowing refugees away) to purchase materials either. He also cannot send others to buy to the abject poverty there and extreme marketplace exclusion. So- on his behalf since they do not know how to purchase quality materi- cial ostracism combines with geographic isolation to reinforce such als at reasonable prices. exclusion. And the individual level cognitive and emotional aspects Social Ostracism. Another element in understanding extreme associated with marketplace literacy further narrow individuals to marketplace exclusion is social ostracism. Social ostracism is asso- the immediate present, and reinforce the inability to overcome ex- ciated with extreme marketplace exclusion as consumers, entrepre- treme marketplace exclusion. In a sense, these are the levers of mar- neurs, and employees. Individuals who fled their country can endure ketplace exclusion. We offer Figure 1 as a theoretical framework to a much more extreme form of ostracism than those in traditional illustrate how these elements relate to each other. subsistence contexts. This element of extreme marketplace exclusion arises from cultural factors, and interestingly, are not confined to national/ethnic differences. Many respondents explain how sellers Advances in Consumer Research (Volume 45) / 935 Discussion Cooper, Elizabeth, and Jo Boyden. “Questioning the Power of Our findings offer contributions to the areas of exclusion, mar- Resilience: Are Children Up to the Task of Disrupting the ketplace exclusion, and subsistence contexts. First, we study mar- Transmission of Poverty?.” (2007). ketplace exclusion in a refugee context where these subsistence Castles, Stephen. “Towards a Sociology of Forced Migration and consumers face the most severe challenges, demonstrating extreme Social Transformation.” Sociology 37, no. 1 (2003): 13-34. marketplace exclusion as a novel phenomenon. Second, we de- Deaton, Angus S. “Wealth Effects on Consumption in a Modified scribe how this phenomenon operates in this context, delineating its Life-Cycle Model.” The Review of Economic Studies 39, no. 4 elements informed by prior research. Third, we illustrate how the (1972): 443-453. elements of extreme marketplace exclusion are interlinked and re- De Montclos, Marc-Antoine Perouse, and Peter Mwangi Kagwanja. inforce each other, creating a vicious cycle of exclusion from the “Refugee Camps or Cities? The Socio-Economic Dynamics of marketplace. In doing so, we integrate aspects of exclusion from pre- the Dadaab and Kakuma Camps in Northern Kenya.” Journal vious studies in refugee contexts and explain it from the marketplace of Refugee studies 13, no. 2 (2000): 205-222. perspective. Dommer, Sara Loughran, Vanitha Swaminathan, and Rohini Ahluwalia. “Using Differentiated Brands to Deflect Exclusion Implications and Protect Inclusion: the Moderating Role of Self-Esteem on Theoretically, we study a qualitatively different level of mar- Attachment to Differentiated Brands.”Journal of Consumer ketplace exclusion and the elements associated with it. Specifically, Research 40, no. 4 (2013): 657-675. abject poverty, geographic isolation, social ostracism, and loss of Doron, Esther. “Working with Lebanese refugees in a Community marketplace literacy are elements leading to extreme marketplace Resilience Model.” Community Development Journal 40, no. 2 exclusion. Even compared with studies of deep exclusion, the severe (2005): 182-191. nature of extreme marketplace exclusion is distinctly unique. De- Duhachek, Adam. “Coping: A Multidimensional, Hierarchical spite potentially having valued skills and education, refugees are of- Framework of Responses to Stressful Consumption Episodes.” ten excluded from the marketplace’s workforce due to the elements Journal of Consumer Research 32, no. 1 (2005): 41-53. we outline. Our work also highlights how individual cognitive and Edwards, Anne. “Working Collaboratively to Build Resilience: A emotional aspects associated with marketplace literacy accentuate CHAT Approach.” Social Policy and Society 6, no. 2 (2007): marketplace exclusion. These individual elements combine with ele- 255. ments rooted in the external environment to lead to the complex phe- Fazel, Mina, and Alan Stein. “Mental Health of Refugee Children: nomenon of extreme marketplace exclusion. Future research should Comparative Study.” Bmj 327, no. 7407 (2003): 134. focus on extreme marketplace exclusion in a variety of contexts to Granovetter, Mark. “The Old and the New Economic Sociology: A better understand how the elements of such exclusion manifest in History and an Agenda.” Beyond the Marketplace: Rethinking the harshest of human conditions, ranging from areas devastated Economy and Society (1990): 89-112. by natural disasters to war. Moreover, future work should examine Haskell, Jane E., and Ashley Storrow. “Using Refugee Voices to the coping strategies consumers employ in such drastic exclusion Improve Cross Cultural Conversations: Research with New circumstances. For example, even when faced with extreme exclu- Mainers.” (2014). sion, our participants still navigate their realities to pursue and attain Hodes, Matthew, and David Goldberg. “The Treatment of belongingness and acceptance. The specific tactics these consumers Refugees.” BJPsych Bulletin (2002): 1-2. leverage to achieve belonging experiences have yet to be closely Korac, Maja. “Integration and How We Facilitate It: A Comparative examined but hold important implications for studying consumers’ Study of the Settlement Experiences of Refugees in Italy and agency and resilience. the Netherlands.” Sociology 37, no. 1 (2003): 51-68. In term of practical implications, we highlight the importance Kuo, Chun-Tung, and Tung-liang Chiang. “The Association of providing refugees with capacity-building learning experiences between Relative Deprivation and Self-Rated Health, that address skills, self-confidence, and awareness of rights. In this Depressive Symptoms, and Smoking Behavior in Taiwan.” regard, the role of marketplace literacy education is particularly rel- Social Science & Medicine 89 (2013): 39-44. evant to extreme marketplace exclusion. Rather than focus on ac- Lee, Edward Ou Jin, and Shari Brotman. “Identity, Refugeeness, cess or financial resources, marketplace literacy education is related Belonging: Experiences of Sexual Minority Refugees in to capacity-building that begins with a deeper understanding of the Canada.” Canadian Review of Sociology/Revue Canadienne de marketplace (know-why) as a basis for know-how and know-what. Sociologie 48, no. 3 (2011): 241-274. Moreover, such education also addresses issues of self-confidence Lee, Jaehoon, and L. J. Shrum. “Conspicuous Consumption and identity, critically important in these settings. In conclusion, our versus Charitable Behavior in Response to Social Exclusion: study of what we term extreme marketplace exclusion provides a A Differential Needs Explanation.”Journal of Consumer foundation to study consumers and employees in the harshest of sub- Research 39, no. 3 (2012): 530-544. sistence contexts, with important theoretical and practical implica- Levitas, Ruth. “The Concept of Social Exclusion and the New tions. Durkheimian Hegemony.” Critical Social Policy 16, no. 46 (1996): 5-20. REFERENCES Levitas, Ruth, Christina Pantazis, Eldin Fahmy, David Gordon, Eva Adkins, Natalie Ross, and Julie L. Ozanne. “The Low Literate Lloyd, and Demy Patsios. “The Multi-Dimensional Analysis Consumer.” Journal of Consumer Research 32, no. 1 (2005): of Social Exclusion.” (2007). 93-105. Loveland, Katherine E., Dirk Smeesters, and Naomi Mandel. “Still Atkinson, Anthony Barnes, and John Hills. “Exclusion, Preoccupied with 1995: The Need to Belong and Preference Employment and Opportunity.” (1998). for Nostalgic Products.” Journal of Consumer Research 37, Betts, Alexander, Louise Bloom, Josiah Kaplan, and Naohiko no. 3 (2010): 393-408. Omata. Refugee Economies: Rethinking Popular Assumptions. University of Oxford, Refugee Studies Centre, 2014. 936 / Extreme Marketplace Exclusion in Subsistence Marketplaces: A Study in a Refugee Settlement in Nakivale, Uganda

Luthar, Suniya S., Dante Cicchetti, and Bronwyn Becker. “The Thompson, Craig J., William B. Locander, and Howard R. Pollio. Construct of Resilience: A Critical Evaluation and Guidelines “Putting Consumer Experience back into Consumer Research: for Future Work.” Child Development 71, no. 3 (2000): 543- The Philosophy and Method of Existential-Phenomenology.” 562. Journal of consumer research 16, no. 2 (1989): 133-146. Mead, Nicole L., Roy F. Baumeister, Tyler F. Stillman, Catherine Troisi, Jordan D., and Shira Gabriel. “Chicken Soup Really Is Good D. Rawn, and Kathleen D. Vohs. “Social Exclusion Causes for the Soul “Comfort Food” Fulfills the Need to Belong.” People to Spend and Consume Strategically in the Service of Psychological Science (2011). Affiliation.”Journal of Consumer Research 37, no. 5 (2011): UNHCR 2004, Basic facts, Who is a Refugee?, viewed 20 January 902-919. 2016. Phillimore, Jenny, and Lisa Goodson. “Problem or opportunity? Viswanathan, Madhubalan, Subsistence Marketplaces, Asylum Seekers, Refugees, Employment and Social Exclusion eBookpartnership, eText, and Stipes Publishing (2013). in deprived Urban Areas.” Urban Studies 43, no. 10 (2006): Viswanathan, Madhubalan, Bottom-Up Enterprise: Insights from 1715-1736. Subsistence Marketplaces, eBookpartnership, eText, and Power, Anne, and William Julius Wilson. “Social Exclusion and the Stipes Publishing (2016). Future of Cities.” (2000). Viswanathan, Madhu, Carlos Torelli, Sukki Yoon, and Hila Riemer. Rose, Damaris. “The Housing Situation of Refugees in Montreal ““Fish Out of Water”: Understanding Decision-Making and Three Years After Arrival: The Case of Asylum Seekers Who Coping Strategies of English as Second Language Consumers Obtained Permanent Residence.” Journal of International through a Situational Literacy Perspective.” Journal of Migration and Integration/Revue de l’integration et de la Consumer Marketing 27, no. 6 (2010): 524-533. migration internationale 2, no. 4 (2001): 493-529. Viswanathan, Madhu, and José Antonio Rosa. “Understanding Sartorius, Kurt, Benn Sartorius, Stephen Tollman, Enid Schatz, Subsistence Marketplaces: Toward Sustainable Consumption Johann Kirsten, and Mark Collinson. “Rural Poverty and Commerce for a Better World.” Journal of Business Dynamics and Refugee Communities in South Africa: A Research 63, no. 6 (2010): 535-537. Spatial–Temporal Model.” Population, space and place 19, Viswanathan, Madhu, José Antonio Rosa, and Julie A. Ruth. no. 1 (2013): 103-123. “Exchanges in Marketing Systems: the Case of Subsistence Schweitzer, Robert, Jaimi Greenslade, and Ashraf Kagee. “Coping Consumer–Merchants in Chennai, India.” Journal of and Resilience in Refugees from the Sudan: A Narrative Marketing 74, no. 3 (2010): 1-17. Account.” Australian and New Zealand Journal of Psychiatry Viswanathan, Madhubalan, Srinivas Sridharan, Roland Gau, and 41, no. 3 (2007): 282-288. Robin Ritchie. “Designing Marketplace Literacy Education in Sen, Amartya. “Social Exclusion: Concept, Application, and Resource-Constrained Contexts: Implications for Public Policy Scrutiny.” (2000). and Marketing.” Journal of Public Policy & Marketing 28, no. Shah, Anuj K., Sendhil Mullainathan, and Eldar Shafir. “Some 1 (2009): 85-94. Consequences of Having Too Little.” Science 338, no. 6107 Walker, Iain, and Heather J. Smith. Relative Deprivation: (2012): 682-685. Specification, Development, and Integration. Cambridge Shah, Anuj K., Eldar Shafir, and Sendhil Mullainathan. “Scarcity University Press, 2002. frames value.” Psychological Science 26, no. 4 (2015): 402- Wang, Jeff Jianfeng, and Qian Tian. “Consumer Vulnerability 412. and Marketplace Exclusion: A Case of Rural Migrants and Sridharan, Srinivas, and Madhu Viswanathan. “Marketing in Financial Services in China.” Journal of Macromarketing 34, Subsistence Marketplaces: Consumption and Entrepreneurship no. 1 (2014): 45-56. in a South Indian Context.” Journal of Consumer Marketing Weidner, Kelly L., José Antonio Rosa, and Madhu Viswanathan. 25, no. 7 (2008): 455-462. “Marketing to Subsistence Consumers: Lessons from Stein, Barry N. “Resettlement of Khmer Refugees in the United Practice.” Journal of Business Research 63, no. 6 (2010): States.” (1982). 559-569. Taylor, Savitri. “Australia’s Safe Third Country’Provisions-Their Werker, Eric. “Refugee Camp Economies.” Journal of Refugee Impact on Australia’s Fulfillment of Its Non-Refoulement Studies 20, no. 3 (2007): 461-480. Obligations.” U. Tas. L. Rev. 15 (1996): 196. White, Judy. Enhancing and Developing Policies, Models and Taylor, Janet. “Refugees and Social Exclusion: What the Literature Practices to Address the Mental Health Needs of Immigrant Says.” (2004). and Refugee Women in Saskatchewan. ProQuest, 2007. Taylor, Janet., and D. Stanovic. “Services and Entitlements: Winlow, Simon, and Steve Hall. Rethinking Social Exclusion: The Refugees, Migrants and Asylum Seekers.” Migration Action End of the Social?. Sage, 2013. 26, no. 2 (2004): 32-35. Large Steps Toward Small Donations: Reputational Benefits of Nominal Corporate Generosity Tiffany Vu, University of Michigan, USA Scott Rick, University of Michigan, USA

EXTENDED ABSTRACT sumers form impressions of companies based on limited informa- One prevalent and increasingly lucrative prosocial practice is tion. In all conditions, we described a clothing retailer with annual for brands to donate some portion of sales or profits to designated profits of $10 million that had conducted a donation campaign over charitable causes via “cause marketing” campaigns (e.g., Krishna, the past year to help raise money for local charities. 2011; Small & Cryder, 2016). Descriptions of such campaigns often We either described the retailer as donating 1% of its profits highlight two attributes: the percent-of-profits (or percent-of-reve- from each purchase, up to a maximum amount of $26,000 (Low- nues) from each purchase that the brand will donate, and the maxi- er%/Higher$ condition) or donating 30% of its profits from each mum amount of money the brand will donate. For example, a 2014 purchase, up to a maximum amount of $25,000 (Higher%/Lower$ campaign by Chipotle locations in Minnesota pledged that “50% of condition). Because we noted that the retailer has annual profits of [its] proceeds up to $35,000 will benefit Como Park Zoo and Con- $10 million, it should be clear to participants that the retailer would servatory.” ultimately donate the maximum amount. We emphasized this fact by We examine how these numerical attributes of cause marketing noting that the retailer “ultimately donated that maximum amount.” campaigns influence perceptions of brands’ generosity. We anticipate Thus, normatively, the retailer should be viewed as (slightly) more that the percent-of-proceeds donated from each purchase will be a generous when they donate $26,000 than when they donate $25,000. stronger predictor of generosity perceptions than maximum dona- We also counterbalanced the order in which the campaign at- tions. Our rationale builds on research by Hsee and colleagues on attributes were presented. Half of the participants were randomly as- tribute evaluability. Hsee and Zhang (2010) proposed that sensitivity signed to view the percent-of-proceeds attribute first (e.g., in the to variation in the value of an attribute (i.e., the attribute’s evaluabil- Lower%/Higher$ condition, the retailer pledged to “donate 1% of its ity) is partly a function of knowledge of the attribute’s distribution profits from each purchase, up to a maximum amount of $26,000”). (e.g., its range or mean). In the cause marketing context, consumers The other half of participants were randomly assigned to view the are unlikely to recall typical values of the percent-of-proceeds and maximum donation attribute first (e.g., in the Lower%/Higher$ con- maximum donation attributes. Nevertheless, consumers are likely to dition, the retailer “pledged to donate up to a maximum amount of be more sensitive to variation in the percent-of-proceeds attribute $26,000, based on 1% of its profits from each purchase”). than variation in maximum donations, because the former attribute We then asked participants to indicate on 0-10 scales (where has a clear upper bound (100%), but the latter attribute does not. 0=not at all and 10=very) the extent to which they found the retailer A novel implication of our reasoning is that cause marketing to be generous. Next, we asked participants two open-ended recall campaigns may allow brands to appear highly generous without ac- questions: “What percentage of profits from each purchase did the tually being highly generous (by donating a high percent-of-proceeds company donate?” and “What was the total amount of money the and a low maximum donation). In what follows, we investigate that company ultimately donated?” Finally, participants provided demo- implication and more broadly examine whether the percent-of-pro- graphic information. ceeds attribute is in fact more predictive of generosity perceptions Results and Discussion than the maximum donation attribute. We explore whether this effect We first conducted a factorial ANOVA treating generosity rat- holds over a range of percentages. To test our evaluability account, ings as the dependent variable and numerical campaign attributes we manipulate the number of different cause marketing campaigns (Lower%/Higher$ vs. Higher%/Lower$) and attribute presentation participants encounter (and thus the evaluability of the campaigns’ order as independent variables. We found a significant main effect numerical attributes). In addition, we attempt to rule out several of numerical campaign attributes (F(1,596) = 12.30, p < .001). As alternative accounts (e.g., that the percent-of-proceeds attribute is predicted, perceived generosity was significantly greater when the more influential because it is the first attribute presented or because company donated a higher percentage of profits and lower maximum it is especially memorable). amount than when the company donated a lower percentage of profits and higher maximum amount (M = 6.69, SD = 2.57 vs. M = 5.93, Experiment 1 SD = 2.73; t(598) = 3.49, p < .001, d = .29). We observed this differ- Our first experiment aimed to document initial evidence of ence despite the fact that participants were assured that the retailers consumers’ greater sensitivity to the percent-of-proceeds attribute of would ultimately donate their maximum amount. Normatively, we cause marketing campaigns than to the maximum donation attribute. would expect the Lower%/Higher$ campaign to be viewed at least We also investigated whether this effect is an artifact of the order in as generous as the Higher%/Lower$ campaign, if not slightly more which the two attributes are presented. Typically, the percent-of-pro- generous. ceeds attribute appears first (e.g., when companies pledge to donate There was no main effect of attribute presentation order X% up to $Y). If people are indeed more sensitive to variation in the (F(1,596) = .12, p = .73), and no interaction between numerical cam- percent-of-proceeds attribute, it could be because they simply paid paign attributes and attribute presentation order (F(1,596) = 1.65, p = greater attention to the first numerical information they encountered. .20). This suggests that greater sensitivity to the percent-of-proceeds We therefore counterbalanced which attribute appeared first. attribute is not an artifact of attribute presentation order. Participants and Procedure We also examined whether the percent-of-proceeds attribute We recruited 600 adults (49% female; mean age: 33) via Ama- was more memorable than the maximum donation attribute. Par- zon Mechanical Turk (MTurk) to participate in exchange for a small ticipants were not significantly more likely to correctly recall the payment. We told participants that we were interested in how con- percent-of-proceeds donated from each purchase than the maximum

Advances in Consumer Research 937 Volume 45, ©2017 938 / Large Steps Toward Small Donations: Reputational Benefits of Nominal Corporate Generosity donation amount (90% vs. 87%; p = .10, Fisher’s Exact Test). If we Higher$ condition (M = 6.53, SD = 2.65 vs. M = 5.13, SD = 2.57; focus only on the 486 participants who correctly recalled both attri- t(302) = 4.66, p < .001, d = .54). Generosity ratings were also signifi- butes, perceived generosity was still greater when the retailer donat- cantly greater in the Higher%+20 condition than in the Lower%+20 ed a higher percentage of profits and lower maximum amount than condition (M = 6.72, SD = 2.70 vs. M = 5.97, SD = 2.70; t(294) = when the retailer donated a lower percentage of profits and higher 2.37, p = .019, d = .28). maximum amount (M = 6.76, SD = 2.57 vs. M = 6.04, SD = 2.71; These patterns suggest our results were not merely driven by t(484) = 3.05, p < .01, d = .27). aversive reactions to a somewhat trivial donation of 1% of profits. Experiment 1 provides initial evidence that, when presented Instead, these results suggest that generosity perceptions are gener- with a description of a cause marketing campaign, consumers’ per- ally more sensitive to the percent-of-proceeds attribute, even when ceptions of the brand’s generosity are more sensitive to the percent- that percentage is not trivially low. of-proceeds to be donated than to the maximum donation amount. As a result, in this experiment, the brand was viewed as significantly Experiment 3 less generous when it was objectively more generous. We found that One implication of our conceptual framework is that providing this was not driven by the order in which the two key attributes were information about other cause marketing campaigns should increase described or differences in the memorability of the two attributes. the evaluability of both the percent-of-proceeds attribute and the maximum donation attribute; it should especially increase the evalu- Experiment 2 ability of the maximum donation attribute (cf. Hsee, 1996). When Our explanation for the greater perceived generosity of the maximum donations become more evaluable, the objectively less Higher%/Lower$ brand is that the percent-of-proceeds attribute is generous campaign should no longer be viewed as more generous. easier to evaluate and thus more influential. However, it is possible Participants and Procedure that the effect may have been driven by reactions to the 1% figure in We recruited 297 adults (49% female, mean age: 33) via MTurk particular, rather the greater evaluability of (all) percentages. That to participate in exchange for a small payment. People who had al- is, the 1% of profits donated in the Lower%/Higher$ condition may ready participated in prior experiments were not eligible to partici- have appeared to be an insultingly trivial gesture. Experiment 2 pate in this experiment. As in previous experiments, we told partici- therefore examined whether generosity perceptions are more sensi- pants that we were interested in how consumers form impressions tive to the percent-of-proceeds attribute even when that percentage of companies based on limited information. In all conditions, we is not trivially low. described a clothing retailer with annual profits of $10 million that Participants and Procedure plans to conduct a donation campaign during the current fiscal year We recruited 600 adults (52% female; mean age: 34) via MTurk to help raise money for local charities. to participate in exchange for a small payment. People who had al- We randomly assigned participants to one of three conditions. ready participated in Experiment 1 were not eligible to participate In two separate evaluation conditions, participants either learned in this experiment. As in Experiment 1, we told participants that we of a cause marketing campaign that donates 1% of its profits from were interested in how consumers form impressions of companies each purchase, up to a maximum amount of $40,000 (SE:Lower%/ based on limited information. In all conditions, we described a cloth- Higher$) or donates 20% of its profits from each purchase, up to a ing retailer with annual profits of $10 million that plans to conduct a maximum amount of $10,000 (SE:Higher%/Lower$). In a third joint donation campaign during the current fiscal year to help raise money evaluation condition (JE), participants viewed both campaigns. The for local charities. two campaigns were presented as the efforts of two different cloth- We randomly assigned participants to one of four conditions. ing retailers. Half of participants were randomly assigned to either a Lower%/ In all three conditions, participants read that all donations Higher$ condition (retailer donates 1% of profits from each - pur would be made at the end of the fiscal year. We included this in- chase, up to a maximum amount of $26,000) or a Higher%/Lower$ formation to rule out the potential interpretation that the Higher%/ condition (retailer donates 25% of profits from each purchase, up Lower$ campaign could reach its target amount faster and donate to a maximum amount of $25,000). We increased these percentages faster than the Lower%/Higher$ campaign. In other words, we did by 20% for the other half of participants. That is, the other half of not want the assumed speed of donation to be confounded with the participants were randomly assigned to a Lower%+20 condition (re- numerical attributes of the campaign. tailer donates 21% of profits from each purchase, up to a maximum We then asked participants to indicate the extent to which they amount of $26,000) or a Higher%+20 condition (retailer donates found the company to be generous on a 0-10 scale, where 0=not at 45% of profits from each purchase, up to a maximum amount of all generous and 10=very generous. Next, we asked participants to $25,000). recall the percent-of-proceeds and maximum donation values from Next, we asked participants to rate the extent to which they the donation campaign(s) they had read about (open-ended ques- viewed the retailer as generous on a 0-10 scale, where 0=not at all tions). As before, the percent-of-proceeds and maximum donation generous and 10=very generous. We also asked participants to re- attributes were about equally likely to be recalled accurately, in both call the percent-of-proceeds and maximum donation values from the separate evaluation and joint evaluation conditions (ps > .45). the campaign they had read about (two open-ended questions). As Results and Discussion in Experiment 1, participants were not significantly more likely to Figure 1 displays perceived generosity by condition. Consistent correctly recall the percent-of-proceeds value than the maximum do- with previous experiments, in the separate evaluation conditions, nation value (90% vs. 87%; p = .11, Fisher’s Exact Test). Finally, perceived generosity was significantly greater in the SE:Higher%/ participants provided demographic information. Lower$ condition than in the SE:Lower%/Higher$ condition (M = Results and Discussion 5.99, SD = 2.82 vs. M = 5.21, SD = 2.59; t(197) = 2.03, p = .044, Consistent with Experiment 1, generosity ratings were signifi- d = .29). However, in the joint evaluation condition, where partici- cantly greater in the Higher%/Lower$ condition than in Lower%/ pants could see both campaigns, the numerical attributes of the cam- Advances in Consumer Research (Volume 45) / 939 paign did not significantly influence perceived generosity (Higher%/ REFERENCES Lower$ M = 6.17, SD = 2.38 vs. M = 6.44, SD = 2.52, t(97) = 1.02, Aaker, J., Garbinsky, E. N., & Vohs, K. D. (2012). Cultivating p = .31). Using the analysis outlined in Hsee (1996, footnote 2), we admiration in brands: Warmth, competence, and landing in found that this was a significant Separate Evaluation/Joint Evaluation the “golden quadrant.” Journal of Consumer Psychology, 22, interaction (t(296) = 2.29, p = .023). Thus, as predicted, we found 191-194. the percent-of-proceeds attribute was more influential only when a Aaker, J., Vohs, K. D., & Mogilner, C. (2010). Non-profits are single cause marketing campaign was evaluated in isolation (which seen as warm and for-profits as competent: Firm stereotypes is likely how consumers encounter cause marketing campaigns in the matter. Journal of Consumer Research, 37(2), 224-237. real world – i.e., one at a time). The percent-of-proceeds attribute is Chansky, E. (2015). For goodness sake: Legal regulation and best less influential when consumers have some comparison information practices in the field of cause marketing. Retrieved from http:// that makes the maximum donation attribute more evaluable. engageforgood.com Chernev, A., & Blair, S. (2015). Doing well by doing good: The General Discussion benevolent halo of corporate social responsibility. Journal of A recent report by Engage for Good (a popular cause marketing Consumer Research, 41(6), 1412-1425. forum; Chansky, 2015) expressed concern that, in cause marketing Desvousges, W. H., Johnson, F. R., Dunford, R. W., Hudson, S. campaigns, a “generous-sounding percentage” of proceeds can be P., Wilson, K. N., & Boyle, K. J. (1993). Measuring natural “misleading” (cf. Olsen, Pracejus, & Brown, 2003) and “naturally, resource damages with contingent valuation. In Contingent what matters, is the actual amount donated.” Our work suggests valuation: A critical assessment (pp. 91-164). Emerald Group. that this concern is well-placed. In three experiments, we found that Hayes, A. F. (2013). An Introduction to mediation, moderation, and brands were viewed as significantly more generous when donating conditional process analysis: A regression-based approach. a higher percent-of-proceeds and a lower maximum amount than New York: Guilford Press. when donating a lower percent-of-proceeds and a higher maximum Hsee, C. K. (1996). The evaluability hypothesis: An explanation for amount. This occurred despite descriptions that made it clear that the preference reversals between joint and separate evaluations of brand would always make its maximum donation. The effect is not alternatives. Organizational Behavior and Human Decision driven by the greater memorability of percentages (Experiment 1) Processes, 67(3), 247-257. and is not limited to cases in which the percent donated is obviously Hsee, C. K., & Zhang, J. (2010). General evaluability low (Experiment 2). The presence of comparative context especially theory. Perspectives on Psychological Science, 5(4), 343-355. helps people evaluate the maximum donation attribute and reduces Krishna, A. (2011). Can supporting a cause decrease donations the undue influence of the percent-of-proceeds attribute (Experiment and happiness? The cause marketing paradox. Journal of 3). Consumer Psychology, 21(3), 338-345. Several open questions remain. The extent to which brands Lin-Healy, F., & Small, D. A. (2012). Cheapened altruism: and retailers intentionally capitalize on the effect documented here Discounting personally affected prosocial actors. is unclear. Some companies have likely benefitted from the effect Organizational Behavior and Human Decision Processes, documented here, whether intentionally or not (e.g., Chipotle’s offer 117(2), 269-274. to donate 50% of sales, up to $35,000). In addition, in many cause Malone, C., & Fiske, S. T. (2013). The human brand: How we marketing campaigns, there will be some uncertainty about whether relate to people, products, and companies. John Wiley & Sons. the brand will reach its maximum donation amount. It would be in- Newman, G. E., & Cain, D. M. (2014). Tainted altruism: When teresting to examine how consumers form expectations about how doing some good is evaluated as worse than doing no good at close brands will get to their maximum. Questions like these seem all. Psychological Science, 25(3), 648-655. worthy of future research. Olsen, G. D., Pracejus, J. W., & Brown, N. R. (2003). When profit equals price: Consumer confusion about donation amounts in cause-related marketing. Journal of Public Policy and Marketing, 22(2), 170-180. Small, D. A., & Cryder, C. (2016). Prosocial consumer behavior. Current Opinion in Psychology, 10, 107-111. More Harm is Less Dangerous Monica Wadhwa, INSEAD, Singapore Mustafa Karatas, Koc University, Turkey

EXTENDED ABSTRACT Four experiments, including two field experiments, examined Health has become a global challenge, with health problems these hypotheses. To provide a strong test of our hypotheses, we increasing across the world. Worldwide obesity rate, for example, compare effectiveness of a message communicating two equally has doubled (NCD-RisC 2016). Flu viruses have become stronger threatening risks with that of a message communicating only one over the years (Savage 2008). Many health problems can be avoided of these two risks. Studies were conducted in Turkey and all health through simple preventive actions. Exercising, for example, could messages were communicated in Turkish. There were no exclusions help prevent obesity. Thus, policy makers are focused on understand- unless otherwise reported. ing how preventive behaviors could be encouraged. A significant part of this focus is on making preventive health messages more effective Experiment 1 (Gerend & Shepherd, 2007). Preventive messages commonly inform Experiment 1, conducted at a health center, examined the pri- the readers of different health risks associated with a health condi- mary hypothesis in the context of obesity—people exposed to two tion. An anti-obesity message, for example, might inform readers of obesity-related risks are less likely to engage in the target preven- obesity-related health risks, such as high blood pressure and diabe- tive action (exercise) and less likely to join an obesity prevention tes. Current research explores whether the number of health risks a program as opposed to those exposed to only one of the two risks. preventive message communicates could affect preventive decision During regular check-ups, as is required, the general physician making. To illustrate, an anti-obesity message could communicate handed over pamphlets describing obesity-related risks to incom- both high blood pressure and diabetes as obesity-related risks or it ing patients (N=240) with a BMI of 25 or over. These pamphlets could communicate one of these two equally threatening and likely included affect-laden pictures of a family on the front page and a health risks. Which of these two strategies would be more effective description of obesity risk(s) on the second page. One version of the in encouraging preventive behavior adoption? pamphlet communicated two risks of obesity—high blood pressure Normatively speaking, a reader should find the health condition and Type 2 diabetes—pretested to be equally threatening and likely more dangerous and thus be more likely to adopt the target preven- (see figure-1 for the sample message). The visual position of the two tive action when the message communicates multiple equally threat- risks was counterbalanced across participants in all experiments. ening and likely health risks (e.g., high blood pressure and diabetes), The other version communicated one of these two risks—high blood compared with when it communicates only one of the two risks. pressure or diabetes. The message encouraged patients to walk regu- Consistently, to encourage adoption of preventive behaviors, policy larly to prevent obesity, an advice typically given by the physician. makers and health experts typically communicate all major risks as- Patients were also asked by the physician to join the obesity preven- sociated with the health condition. tion program run by the health ministry at the center. In contrast to the normative view, we propose that presenting One week after each patient’s visit, a nurse contacted the patient multiple equally threatening health risks is a less effective strategy and measured the frequency as well as the duration of the patient’s than presenting only one of these risks—a proposition drawn upon walks during the week. This nurse also recorded whether the patient research on imagery. Research in this domain suggests that one’s ca- had joined the obesity clinic—this information was checked against pacity to simulate mental images is limited (Baddeley & Andrade, the clinic records. 2000). Drawing upon this research on imagery, current research is Results and Discussion the first to argue that due to mental imagery capacity limitations, We used a two-part model to examine the impact of the number health risks should be less vividly imagined when one has to men- of risks on walking duration. The two-part model analysis revealed tally simulate multiple health risks simultaneously, compared with that those in the two-risk condition were less likely to walk (30.83%) when one has to simulate only one health risk. than those in the one-risk condition (47.5%; z=-2.63, p=.009, B= Vivid imagery has been shown to be an important determinant - .707, CI [- 1.23, -.180]). Amongst those who walked, those in the of people’s behaviors, such that a message that evokes more vivid two-risks condition (M=33.15 minutes; SD: 17.13) walked less than mental images is likely to be more effective in persuading behaviors those in the one-risk condition (M=44.49 minutes; SD: 27.55; z=- (Nisbett & Ross, 1980). Drawing upon a synthesis of these findings, 2.24; p=.025; CI [- 21.275, -1.396]; d=.49). we propose that a health message communicating multiple equally Finally, a binary logistic regression revealed that patients were threatening health risks would lead to less vivid imagery and, there- less likely to join the obesity program when informed of two risks fore, be less effective in encouraging adoption of preventive behav- (3.33%) compared with when exposed to one risk (10.83%; Wald χ2 iors, compared with a message that communicates only one of the = 4.599; p=.032; B= -1.259, Odds Ratio=.284; CI[.090, .897]). two health risks. This study provides support for our primary hypothesis: Com- Research suggests that when people use affective processing, municating multiple health risks in a preventive message reduced the vivid imagery is an important determinant of their behaviors (Keller likelihood of engaging in preventive behaviors than presenting only & McGill, 1994). However, when people use a deliberative mode one of the multiple risks. of processing, they tend to rely on logical, calculation-based reactions (Hsee & Rottenstreich, 2004). Therefore, we further propose Experiment 2 that when a deliberative mindset is induced, people should be more Experiment 2, conducted at a pharmaceutical store, sought to sensitive to the number of health risks. Thus, when people are in replicate the effect found in previous study in the context of flu. a deliberative mindset, communicating multiple health risks in the Customers (N=432) at the store were informed of flu-related message should be a more effective in encouraging adoption of pre- risks using posters, which were posted at the entrance and at the ventive behaviors than presenting only one of these risks. check-out counter. These posters communicated either two equal-

Advances in Consumer Research 940 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 941 ly threatening and likely flu-related risks—muscle aches and sore Experiment 4 throat, or only one of these two risks—and encouraged customers In the last experiment, we further examined our underlying to keep their hands clean to prevent flu. Additionally, the posters re- conceptualization by examining the mediating role of vivid imag- ferred to Flu as Grip-Nezle, an affective term used by lay people to ery. We predicted that when relying on affective reactions, preven- refer to flu in Turkey. In order to examine whether presenting two tive message should evoke less vivid images when it communicates risks would be as good as sharing a generic health message, in a third two health risks as opposed to one of the two health risks. However, condition, a neutral poster communicating a generic health message when participants are cued to focus on their deliberative reactions, was posted. Hand-sanitizing wipes packets were available for sale at number of health risks should not impact vivid imagery. Moreover, the check-out counter; sale of wipes was our main dependent vari- we expected vivid imagery to mediate the impact of number of able. health risks on preventive action in the affective-cue condition, but Results and Discussion not in the deliberative-cue condition. To examine the aforementioned hypotheses, during flu season, We ran a logistic regression with two contrasts. The first con- participants (N=146) were exposed to a preventive message encour- trast showed that there was no difference in the likelihood to purchase aging them to keep their hands clean to prevent flu. This message hand-sanitizing wipes between those exposed to two-risks (6.16%) communicated either two equally threatening and likely flu-related and those exposed to neutral health message (5.59%; p=.837).The risks—muscle aches and sore throat, or only one of these two risks. second contrast showed that in line with the findings of the previ- Drawing upon research (Sinaceur, Heath & Cole 2005), which ous study, those exposed to a single risk were more likely to buy shows that using an affective (vs. scientific) frame to refer to a dis- the hand-sanitizing wipes (13.29%) than those exposed to two risks ease can induce affective (vs. deliberative) processing mode, we var- or neutral message (5.88%; Wald χ2 = 6.541; p=.011; B=.908; Odds ied the frame (affective versus scientific) used to refer to flu. One Ratio=2.478; CI[1.236, 4.968]). Findings of Experiment 2 replicated set of messages used in the current study referred to flu as Grip- the effect found in the previous study in the context of flu. Nezle, an affective term used for flu by laypeople in Turkey (affective Given the previous experiments employed affective pictures or frame). Another set of messages (scientific-frame) referred to flu asA affective terms, we assume that participants relied on their affective Tipi Enflüanza (Influenza-Type A), a term used by doctors in Turkey. reactions while making the decision. However, the previous experi- These terms were chosen based on a pretest. ments do not explicitly test this assumption. The next two experi- After evaluating the message, participants chose a gift between ments take care of this limitation. two items—a pen and a hand-sanitizing wipes pack, which served as Experiment 3 the dependent measure. Participants then responded to a final survey, Experiment 3 examines the moderating role of affective versus which asked them to evaluate the viewed message on six nine-point deliberative processing. We argue that when cued to rely on af- scale items (Keller & Block 1997) measuring how vividly they could fective reactions, we should replicate the effect found in previous imagine the message. Four participants did not take this survey, leav- studies. When cued to rely on more deliberative reactions, the effect ing us with a sample of 142 participants for these measures. should be reversed such that participants should be more likely to Results and Discussion engage in the preventive action when exposed to multiple risks as A logistic regression revealed a significant interaction between opposed to only one risk. the frame-type and number of risks (z=3.9615; p=.0001; CI[1.4346, Participants (N=273) first engaged in a task, which cued them to 4.2442]). In the affective-frame condition, fewer participants chose focus on either their affective or deliberative reactions. Drawing upon hand-sanitizing wipes when informed of two risks (36.11%) com- past research (Hsee& Rottenstreich, 2004), those in the affective-cue pared with when informed of one of the two risks (75.75%; z = condition responded to questions that required them to examine and -3.2008; p=.0014; B=-1.71; CI[- 2.757,-.663]; Odds Ratio=.1808). report their feelings toward different affect-laden words. Participants In the scientific-frame condition, participants were more likely to in the deliberative-cue condition solved simple math problems. Sub- choose hand-sanitizing wipes when two flu-related risks were com- sequently, participants viewed one of the two versions of an anti- municated (58%) than when one risk was communicated (30%; z = obesity message encouraging them to eat health. As in Experiments 2.363; p=.018; B=1.129; CI[.193, 2.066] Exp(B)=3.092). 1 and 3, one version communicated two risks of obesity—high blood An ANOVA conducted on the vivid imagery measure revealed pressure and diabetes. The other versions communicated only one of a significant interaction between number of risks and frame-type these two risks—high blood pressure or diabetes. Participants then factors (F(1,138)=29.63, p < 0.001). In the affective-frame condi- chose a free snack between an apple and a brownie. tion, images elicited by the message were less vivid when the mes- A logistic regression revealed a significant type of cue by num- sage communicated two risks (M=3.143; SD=1.379) than when it ber of risks interaction (z=4.4134; p<.0001; CI[1.3490, 3.5043]) . In communicated one risk (M=4.887; SD=1.163; F(1,138)=35.15, p the affective-cue condition, participants were less likely to choose an < 0.001; CI[-2.326, -1.163,]; d=1.36). No such corresponding dif- apple when informed of two risks (28.89%), compared with when ference was found in the scientific-frame condition (F(1,138)=2.62, informed of one of the two risks (60%; z=-3.3236; p=.0009; B=- p>.1; CI[-.098, .984]). We further conducted a moderated mediation 1.3063; CI[-2.0766, -.5359]; Odds Ratio=.2708). In contrast, in the using PROCESS model 7 (Hayes 2013), which revealed that vivid deliberative-cue condition, participants were more likely to choose imagery mediated the impact of number of risks on likelihood to an apple when informed of two risks (68.89%), compared with when choose hand-sanitizing wet wipes in the affective-frame condition informed of only one risk (41.93%; z=2.9137; p=.0036; B=1.1204; (B=-1.3034, CI=[-2.1902, -0.6375]) but not in the scientific-frame CI[.3667, 1.874]; Odds Ratio=3.066). condition (B=0.3310, CI=[-0.0398, 0.8615]). Providing support for the underlying conceptualization, the cur- This experiment provides further support for our conceptual- rent study shows that increasing the number of health risks commu- ization related to vivid imagery. In the affective frame condition, nicated in the message can negatively (positively) impact the likeli- presenting two health risks elicited weaker images of the message, hood of adopting the target preventive action when one is cued to which mediated the impact of the number of risks on the likelihood rely on affective (deliberative) reactions. of adopting preventive behavior. However, there was no such cor- 942 / More Harm is Less Dangerous responding difference in vivid imagery in the scientific frame condi- REFERENCES tion. Baddeley, Alan D. and Jackie Andrade (2000), “Working Memory and the Vividness of Imagery,” Journal of Experimental General Discussion Psychology: General, 129(1), 126-45. Current research is the first to reveal that people’s likelihood of Gerend, Mary A., and Janet E. Shepherd (2007), “Using Message adopting a preventive action non-normatively decreases when the Framing to Promote Acceptance of the Human Papillomavirus health message communicates two health risks associated with the Vaccine,” Health Psychology, 26 (6), 745-52. condition as opposed to one of the risks. Our findings further show Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, that when nudged to rely on deliberative reactions, people focus on and Conditional Process Analysis: A Regression-Based the number of health risks, thus increasing the likelihood to adopt a Approach, Guilford Press. preventive action when the message communicates multiple risks as Hsee, Chris K., & Yuval Rottenstreich (2004), “Music, pandas, and opposed to only one risk. muggers: on the affective psychology of value,”Journal of These findings make important theoretical contributions—First, Experimental Psychology: General, 133(1), 23. our findings contribute to the affective decision making research. Keller, Punam Anand, and Lauren G. Block. “Vividness effects: This is the first research to show that in the context of negative out- A resource-matching perspective.” Journal of Consumer comes, people are less likely to adopt a preventive action when it Research 24.3 (1997): 295-304. can potentially prevent multiple equally threatening health risks as Larrick, Richard P. “Motivational factors in decision theories: The opposed to a single health risk. Second, our findings contribute to role of self-protection.” Psychological Bulletin 113.3 (1993): research on vivid imagery. Our findings show that increasing the 440. number of negative outcomes communicated in a message can re- NCD Risk Factor Collaboration (NCD-RisC) (2016), “Trends in duce the vividness with which one can imagine the risks, but only Adult Body-Mass Index in 200 countries from 1975 to 2014: when relying on affective reactions, not when one is cued to rely on A Pooled analysis of 1698 Population-based measurement deliberative reactions. studies with 19.2 million participants,” The Lancet, We believe our findings have important policy making implica- 387(10026), 1377-1396. tions. Public health campaigns often tend to communicate all pri- Nisbett, Richard E. and Lee Ross. (1980). Assigning Weights to mary risks associated with a health condition. Given a significant Data: The “Vividness Criterion.” In R. Nisbett & L. Ross amount of decision making is driven by people’s intuitive, affective (Eds.), Human inference: Strategies and shortcomings of reactions (Larrick 1993), current research findings suggest that com- social judgment (pp. 43–62). Englewood Cliffs, NJ: Prentice- municating multiple risks is a strategy which can unintentionally Hall. reduce the likelihood of adopting preventive behaviors. Our find- Savage, S. (2008, May 20). Flu getting stronger, more resistant. ings further suggest that for experts, who tend to rely on deliberative Red Orbit. Retrieved from http://www.redorbit.com/news/ cues, presenting multiple health risks is likely to be a more effective health/1395429/flu_getting_stronger_more_resistant strategy than presenting a single health risk. In essence, our findings Sinaceur, Marwan, Chip Heath, and Steve Cole (2005), “Emotional present insights into how health communication could be made more and Deliberative Reactions to a Public Crisis: Mad Cow effective depending on the audience. Disease in France,” Psychological Science, 16 (3), 247-54. When Numbers Make You Feel: The Impact of Round versus Precise Numbers on Preventive Health Behavior Monica Wadhwa, INSEAD, Singapore Kuangjie Zhang, Nanyang Technological University, Singapore

EXTENDED ABSTRACT Results and Discussion Prevention is better than cure—this adage has shaped the US We measured the amount of unhealthy eating in the follow- healthcare spending, with about 75% of healthcare expenditure spent ing two ways. First, a linear regression analysis conducted on the on preventive care (Begley 2013). Many health problems can be number of unhealthy food items participants reportedly ate revealed avoided through simple preventive actions. Policy makers are, there- that those exposed to round numbers consumed significantly less un- fore, keen on understanding how preventive health messages can be healthy food items (M=1.90) than those exposed to precise numbers made more persuasive (Gerend and Shepherd 2007). To make mes- (M=2.65; t(154) = -2.91, p=.004). While, these participants also re- sages more persuasive, preventive health messages often incorporate ported consuming a slightly higher number of healthy food items numerical cues, such as the probability with which a preventive ac- (M=4.65), than those in the precise number condition (M=4.08), this tion can reduce a health risk (Reyna et al. 2009). The current research difference was not significant (t(154)=1.30, p=.20). makes a novel prediction that incorporating numerical cues that ap- Second, to further examine if round versus precise numbers pear round (e.g., 15.00%) versus precise (e.g., 15.23%) can increase impacted unhealthy eating behaviors, we recruited 202 independent people’s likelihood of adopting preventive behaviors. coders from the same population to rate the food items. Each coder Research shows that round numbers are more easily processed was assigned to rate the food items listed by 15 to 16 participants on than precise numbers, in part because people are more frequently an 8-point scale (1=very unhealthy, 8=very healthy). We computed exposed to round numbers in everyday language (Kettle and Häubl an unhealthy eating score (reverse-coded) by averaging the ratings 2009; Wadhwa and Zhang 2015). Moreover, research on affective of all the food items each participant had consumed. An ANOVA decision-making suggests that exposure to fluent stimuli is likely to conducted on this score revealed that those exposed to round num- engage the default affective system, leading people to rely on their bers had a lower unhealthy eating score (M=3.97) than those exposed intuitive, affective reactions while making decisions (Alter et al. to precise numbers (M=4.28, t(154)=-2.29, p=.02). Study 1A shows 2007). Drawing upon this logic, we propose that exposure to round that presenting obesity-related risks using round versus precise num- numbers should lead people to rely on their more intuitive, affective bers in a health message reduced participants’ unhealthy food con- reactions, compared with exposure to precise numbers. sumption. Following from the idea that stronger affective reactions can motivate preventive health behaviors (Loewenstein et al. 2001; Sina- Study 1B ceur, Heath, and Cole 2005), we further hypothesize that when nu- This study, conducted in two stages, sought to conceptually rep- merical cues in preventive messages are presented in a round (e.g., licate study 1A in the context of water consumption. In the first stage, 15.00%) versus a precise (e.g., 15.23%) format, people are more 260 US participants were presented with a health message featuring likely to adopt preventive behaviors. We examine this hypothesis a man and a woman suffering from flu. The only difference between across five studies. conditions being the probability numbers:

Study 1A 20.00% (vs. 20.37%) of the US population suffered from flu last Study 1A examines our primary hypothesis—round numerical year. cues should increase preventive behavioral intention, compared to precise numerical cues. The number of people being hospitalized due to flu complication This study was conducted in two stages. In the first stage, 200 has increased by 40.00% (vs. 40.21%). US participants were presented with a health message featuring an Drinking water frequently can reduce the risk of getting flu. obese person. The only difference between conditions being the probability numbers (see below): Drink More Water! 30.00% (vs. 31.57%) of the US population suffer from obesity today. Twenty-four hours later, participants were emailed the second survey. Out of the 260 participants recruited in the first stage, 184 15.00% (vs. 15.29%) of deaths in the US were caused by obesity. (45.1% women, Mage= 33.9) completed the second survey. In this second survey, participants reported the number of glasses of water and Eat Healthy! the number of times they had consumed water in the last 24 hours on Twenty-four hours after being exposed to the message, partici- a 21-point scale (ranging from 1 to 20+; 20+ was coded as 21). Final- pants were emailed the second survey. 156 participants (48.1% wom- ly, participants indicated how many glasses of water and how many times they generally consume water in one day (excluding the last 24 en, Mage=31.7) completed the second survey. In this second survey, participants first reported the number of unhealthy, as well as healthy hours) using the same 21-point scale items (ranging from 1 to 20+). food items they had consumed in the past 24 hours. Participants then Our main dependent variable was the change in water consumption listed all the items with approximate portion size they had consumed from the regular water consumption, in 24-hours subsequent to mes- in the last 24 hours. Our main dependent variable was the amount of sage exposure. unhealthy eating participants had engaged in 24 hours, subsequent to Results and Discussion the preventive health message exposure. We first computed the change in thenumber of glasses of water participants drank by subtracting the number of glasses of water they

Advances in Consumer Research 943 Volume 45, ©2017 944 / When Numbers Make You Feel: The Impact of Round versus Precise Numbers on Preventive Health Behavior usually consumed from the number of glasses they drank subsequent the right direction (Lee, Amir, and Ariely 2009). Affective reactions to being exposed to the preventive message. A linear regression re- are more likely to drive behaviors when people believe trusting their vealed that those exposed to round numbers showed a significant in- affective reactions would help them make the right decision. In this crease in the number of glasses of water they drank (M=0.55), com- study, therefore, we explicitly manipulated participants’ situational pared with those exposed to precise numbers (M=-0.01; t(182)=2.23, trust in their feelings. We predicted that when participants’ trust in p=.03). Further, one-sample t tests confirmed that the increase (0.55) feelings is high, exposure to round numbers in a preventive message in number of glasses of water consumed in the round number con- should increase the likelihood to engage in preventive behaviors. In dition was significantly different from zero (p=.002), whereas the contrast, when participants’ trust in feelings is low, impact of round change (-0.01) in the precise number condition was not (p=.95). numbers on preventive behavioral intentions should get attenuated. Similarly, a linear regression revealed that participants in the The study followed a 2 (number: round vs. precise) × 2 (trust round number condition showed a significant increase in the num- in feelings: high vs. low) between-subjects design. 345 US partici- ber of times they drank water (M=0.68) than those in the precise pants (47.0% women, Mage=34.8) completed the study. One partici- number condition (M = -0.29; t(182)=2.73, p=.007). One-sample t pant reported being below age 18 and thus data from this participant test confirmed that the increase (0.68) of number of times water was was removed prior to any analyses. Another participant who took consumed in the round number condition was significantly different approximately 24 hours to complete the experiment was removed from zero (p=.007), whereas the change (-0.29) in the precise num- from further analyses. A post-hoc check revealed that including these ber condition was not (p=.26). Study 1B provides further support for two participants’ data do not change our results. the round number hypothesis in the context of flu. Participants were informed that the first part of the experiment sought to understand how people use feelings when making a deci- Study 2 sion. In this task, participants were asked to recall and describe either We measured actual behavior in the current study. Participants a situation in which they followed their feelings and it was the right were exposed to a dental hygiene message, which encouraged them thing to do (high-trust in feelings) or a situation in which they fol- to floss to avoid gum diseases. We measured the time participants lowed their feelings and it was the wrong thing to do (low-trust in spent on flossing. 202 participants recruited from a major university feelings; adapted from Lee et al. 2009). Fifty-seven participants did in Singapore (62.4% women, Mage=21.3) were randomly assigned to not indicate any experience, and thus data from these participants either the round or the precise number condition. Participants were could not be used for analyses. Additionally, five other participants shown a message related to gum diseases. This message, which in- responded to the question in non-English languages and thus were corporated a gender-neutral picture of an individual suffering from removed prior to any analyses. Removing these participants left us bleeding gum, indicated gum disease related risk estimates as either with a total sample of 286 participants. precise or round numbers: Subsequently, in a purportedly unrelated study, participants read the following scenario (adapted from Sinaceur et al. 2005) with 60.00% (vs. 60.41%) of people in Singapore suffer from gum the only difference across conditions being the probability number: diseases leading to swollen and painful gums. You have just finished eating your dinner. While watching the evening news on TV, you find out that eating chicken may expose Gum diseases could increase the risk of losing teeth by 40.00% (vs. you to the human variant of Bird Flu. According to the recent report, 40.37%). one type of vaccination was recently launched in the US, which can reduce the chance of getting this disease by 60.00% (vs. 60.41%). Flossing Helps Prevent Gum Diseases . Participants were asked to indicate their intention to take the Subsequently, participants were given an opportunity to floss vaccination on a nine-point scale (1=very unlikely, 9=very likely). their teeth. Participants were asked to pick up envelops carrying Results and Discussion individual floss sticks, a paper glass and napkins. Unbeknownst to An ANOVA revealed a significant two-way interaction between the participants, we recorded the time participants took for flossing. the number exposure and the trust in feelings (F(1, 282)=5.44, p=.02; Two participants reported not being able to floss due to their braces. see Figure 1) factors. Follow up analyses reveal that in the high-trust Another ten participants did not pick up the flossing materials and in feelings condition, participants indicated a higher intention to take missed this part of the task. Data from these 12 participants could not the vaccination when the numerical cues presented in the message be used, leaving us with a total sample of 190 participants. were round (M=5.56), than when those cues were precise (M=4.68; Results and Discussion t(282)=2.15, p=.03). However, in the low-trust in feelings condition, Due to the large variation in the flossing time measure, we first this effect of number on the intention to take the vaccination was examined the data for outliers. Based on the three SD above or below eliminated (Mround = 4.90 vs. Mprecise=5.38; t(282)= -1.33, p=.25). the mean criterion (McClelland 2000), we identified and removed an outlier (>3.5 SD above the mean) prior to further analyses. A linear Study 4 regression revealed that participants spent longer time flossing their This study examines the reliance on affective reactions concep- teeth after reading the gum disease message with round numbers tualization in two ways: 1) We argue that if round versus precise (M=58.38 seconds), compared with those who read the gum dis- numbers lead people to focus on their affective reactions, then in- ease message with precise numbers (M=46.58 seconds; t(187)=2.02, cidental exposure to round versus precise numbers in an unrelated p=.045). This study replicated the round number effect using actual task could also subsequently increase people’s preventive behavioral behavioral measures. intention and 2) we examine the role of reliance on affective reactions in moderating this effect. In this study, 483 US participants (60.2% women, M =35.3) Study 3 age The degree to which affective reactions can drive behaviors de- were randomly assigned to one of the two number conditions. Par- pends on whether people trust their feelings would direct them in ticipants engaged in eight consecutive trials of a number sorting task, which asked them to sort four different images of numbers in an as- Advances in Consumer Research (Volume 45) / 945 cending order of the physical size of the images. Participants sorted REFERENCES either round number images (e.g., 10.00, 60.00) or precise number Alter, Adam L., Daniel M. Oppenheimer, Nicholas Epley, images (e.g., 10.32, 63.74). The time participants spent on the task and Rebecca N. Eyre (2007), “Overcoming Intuition: did not differ between conditions (F<1, p=.38). A pretest with 42 Metacognitive Difficulty Activates Analytic Reasoning,” participants indicated that the image sorting task was perceived to Journal of Experimental Psychology: General, 136 (4), 569- be equally difficult (Mround=2.17 vs. Mprecise=1.85, F<1, p=.37) and 76. effortful (Mround=4.32 vs. Mprecise=4.43, F<1, p=.84) between the two Begley, Sharon (2013), Think Preventive Medicine Will conditions. Save Money? Think Again, Retrieved from http://www. Subsequently, participants read a purportedly unrelated preven- reuters.com/article/2013/01/29/us-preventive-economics- tive message featuring a picture of an obese person, accompanied by idUSBRE90S05M20130129. the following message: “Increasing numbers of the US adults suffer Gerend, Mary A., and Janet E. Shepherd (2007), “Using Message from obesity today. Eat Healthy!” Framing to Promote Acceptance of the Human Papillomavirus Participants then indicated their likelihood of avoiding eating Vaccine,” Health Psychology, 26 (6), 745-52. anything unhealthy for dinner that day (1=very unlikely, 9=very Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, likely). We then measured negative affective reactions by asking par- and Conditional Process Analysis: A Regression-Based ticipants two questions: “How worried do you feel about becoming Approach, Guilford Press. overweight or obese?” (1=not at all worried, 9=very worried) and Kettle, Keri, and Gerald Häubl (2010), “Numeric Fluency and “How concerned do you feel about becoming overweight or obese?” Preference,” In Campbell, M. C., Inman, J., & Pieters, R. (1=not at all concerned, 9=very concerned; α=.96). (Eds.) Advances in Consumer Research, Vol. 37, Duluth, MN: Results and Discussion Association for Consumer Research. Lee, Leonard, On Amir, and Dan Ariely (2009), “In Search of First, a linear regression analysis revealed that participants were Homo Economicus: Cognitive Noise and the Role of Emotion more likely to avoid eating unhealthy after being exposed to round in Preference Consistency,” Journal of Consumer Research, 36 (M=5.89) versus precise (M=5.43; t(481)=2.24, p=.026) numbers. (2), 173-87. Next, we examined if number condition impacted the degree Loewenstein, George F., Elke U. Weber, Christopher K. Hsee, and to which affective reactions predicted intention to avoid unhealthy Ned Welch (2001), “Risk as Feelings,” Psychological Bulletin, eating. In order to do so, we computed an interaction term between 127 (2), 267-86. number factor and affective reactions. Our analysis (process model McClelland, Gary H. (2000), “Nasty Data,” Handbook of Research 1; Hayes 2013) revealed a significant two-way interaction between Methods in Social Psychology, 393-411. the number exposure manipulation and affective reaction (b=.21, Reyna, Valerie F., Wendy L. Nelson, Paul K. Han, and Nathan t(479)=2.51, p=.01). In the round number condition, negative affec- F. Dieckmann (2009), “How Numeracy Influences tive reactions significantly predicted intention to avoid unhealthy Risk Comprehension and Medical Decision Making,” eating (β=0.26, t(479)=4.45, p<.0001). However, in the precise num- Psychological Bulletin, 135 (6), 943-73. ber condition, negative affective reactions did not predict intention to Sinaceur, Marwan, Chip Heath, and Steve Cole (2005), “Emotional avoid unhealthy eating (β=0.05, t(479)=0.86, p=.39). and Deliberative Reactions to a Public Crisis: Mad Cow General Discussion Disease in France,” Psychological Science, 16 (3), 247-54. Simple preventive actions, such as getting immunized, avoiding Wadhwa, Monica, and Kuangjie Zhang (2015), “This Number Just unhealthy eating etc., could prevent many health problems and their Feels Right: The Impact of Roundedness of Price Numbers on associated costs. Across different settings using both intentional and Product Evaluations,” Journal of Consumer Research, 41 (5), behavioral measures, this research shows that exposure to round ver- 1172-85. sus precise numbers can lead people to respond more affectively and thereby increase people’s likelihood to adopt preventive behaviors. These findings contribute both to literature on numerical cognition and affective decision making. Access to health-related information has never been easier. Whether it is understanding more about effectiveness of different medical treatments or the risks associated with getting a disease, people can now easily access information through online and offline sources. Much of this information is presented numerically. Our findings show that a simple strategy of expressing such numerical information using round (vs. precise) numbers can increase adoption of preventive behaviors, thereby reducing health risks and ultimately saving lives. 946 / Range Goals as Dual Reference Points: Insights for Effective Goal-Setting and Pursuit Range Goals as Dual Reference Points: Insights for Effective Goal-Setting and Pursuit Scott G. Wallace, Duke University, USA Jordan Etkin, Duke University, USA

EXTENDED ABSTRACT ing’ strategy perform significantly better. Thus the lower aggregate Goal-setting is a crucial determinant of performance and well- performance of range (vs. specific) goal pursuers, as observed in our being. But how should goals be set? The prevailing answer focuses data and in prior research, is driven by a relatively small number of on the specificity and difficulty of a goal’s objective (Locke and individuals who adopt an ineffective strategy for pursuing their range Latham 1990, 2002). Specific, challenging goals generally improve goal (i.e., ‘select lower,’ 18% in each study). performance relative to non-specific “do-your-best” goals, and To test the causal effect of reference-point strategy on perfor- “moderately vague” goals like range goals (e.g., lose 10-15 pounds) mance, Study 3 manipulates strategies using instructions to encour- fall in between. Specific goals are thus widely considered the “gold age either a ‘select upper’ or a ‘switching’ strategy. Range goal pursu- standard” of effective goal-setting. ers who received the ‘switching’ instructions performed significantly We propose an alternative approach to understanding range better than those who received the ‘select upper’ instructions. This goals. Drawing on the theory of goals-as-reference points (Heath, demonstrates the causal effect of the proposed ‘switching’ strategy as Larrick, and Wu 1999), we construe range goals as offering two pos- a way to enhance motivation by leveraging both range endpoints as sible reference points (vs. the single reference point of specific goals; reference points. Furthermore, it indicates the possibility of improv- Scott and Nowlis 2013), and posit that these dual reference points ing range goal performance through simple instructions to encourage can be leveraged to improve performance. Extending prior work the adoption of more effective strategies. treating range goals as less precise (and thus less motivating) than From earning a promotion to losing weight to saving for retire- specific goals (Locke and Latham 1990; Wright and Kacmar 1994), ment, effective goal-setting is crucial to obtaining desired outcomes. we demonstrate that range goals can sometimes be more motivating While the notion that specific goals improve performance is widely than specific goals. accepted, the current research shows that when the dual endpoints of When pursuing a range goal, people can adopt one of three a range goal are treated as sequential reference points, range goals reference-point strategies: focusing on just the lower endpoint, just can outperform specific goals set at the top of the range. This insight the upper endpoint, or switching between the lower and upper end- can help consumers succeed at a variety of personal goals, and can points. The lower endpoint of the range defines the outcome that also help to enhance motivation and improve outcomes in group goal avoids failure, so intuitively, some people may adopt it as their (sole) pursuits (e.g., charitable fundraising), marketing initiatives (e.g., loy- reference point. The upper endpoint, in contrast, demarcates a high alty reward programs), and organizational or educational settings. level of performance, so more motivated individuals may adopt it as Our findings contribute to understanding how multiple - refer their (sole) reference point (treating it like a high specific goal). Al- ence points influence judgment and decision-making. Although sev- ternatively, some people may use both range endpoints as reference eral disciplines have studied the effects of goals-as-reference points points by, for example, initially focusing on the lower endpoint (e.g., (Abeler, Falk, Goette, and Huffman 2011; Allen, Dechow, Pope, and lose 10 pounds) then switching to the upper endpoint (e.g., lose 15 Wu 2016; Dai, Milkman, and Riis 2014, 2015; Heath et al. 1999; pounds) after accumulating goal progress. Pope and Simonsohn 2011), how people choose (or switch) between We propose that treating the dual endpoints of a range goal as multiple available reference points is less well understood. A few sequential targets and switching from the lower to the upper end- investigations have included both a specific goal and a baseline or point produces the highest level of performance—even compared to starting point (Bonezzi, Brendl, and De Angelis 2011; Koo and Fish- a specific goal set at the top end of the range. A key tenet of goals-as- bach 2008; March and Shapira 1992), but none consider the dual reference-points theory is that motivation increases with proximity endpoints of a range goal. By identifying strategies people adopt for to a salient reference point (Heath et al. 1999; Kivetz, Urminsky, range goals (and consequences for performance), the present investi- and Zheng 2006). Relative to focusing on one endpoint or setting gation furthers understanding of this important question. an equivalent specific goal, switching from the lower to the upper Why do some people focus on just one endpoint of the range endpoint should keep goal pursuers closer to their salient reference rather than switch? Properties of the task (e.g., difficulty, enjoyment) point for more of the goal pursuit process, improving motivation and or the range goal itself (e.g., width, incentives, being self-set vs. as- performance. Thus while high specific goals may be the gold stan- signed) might influence what reference-point strategy people adopt, dard in the aggregate, range goals can match (‘select upper’) or even but this merits further investigation. Examining these and other fac- exceed (‘switching’) the performance of specific goals, depending on tors that shape range goal pursuers’ strategies can allow for further the reference-point strategy used. refinement of the present findings, which offer valuable insights for Three experiments test these predictions in the context of an effectively setting and pursuing range goals. effortful proofreading task, identifying what strategies consumers naturally adopt when pursuing range goals (“find 8-12 errors”) and REFERENCES how those strategies affect performance. Study 1 identifies- refer Abeler, J., Falk, A., Goette, L., & Huffman, D. (2011). Reference ence-point strategy using a funneled debrief and Study 2 measures points and effort provision. The American Economic reference point focus (lower or upper endpoint) repeatedly over Review, 101(2), 470-492. the course of the task. In both cases, results show that the proposed Allen, E.J., Dechow, P.M., Pope, D.G., & Wu, G. (2016). ‘switching’ strategy (31% of participants in Study 1, 45% in Study 2) Reference-dependent preferences: Evidence from marathon and the ‘select upper’ strategy (32% in Study 1, 26% in Study 2) are runners. Management Science, forthcoming. most common. Furthermore, as predicted, range goal pursuers who Bonezzi, A., Brendl, C. M., & De Angelis, M. (2011). Stuck in the use the ‘select upper’ strategy perform no worse than those with a Middle The Psychophysics of Goal Pursuit. Psychological high specific goal (“find 12 errors”) and those who use the ‘switch- science, 22(5), 607-612. Advances in Consumer Research (Volume 45) / 947

Dai, H., Milkman, K. L., & Riis, J. (2014). The fresh start ______(2002). Building a practically useful effect: Temporal landmarks motivate aspirational theory of goal setting and task motivation: A 35-year behavior. Management Science, 60(10), 2563-2582. odyssey. American psychologist, 57(9), 705. Dai, H., Milkman, K. L., & Riis, J. (2015). Put your imperfections March, J.G., & Shapira, Z. (1992). Variable risk preferences and the behind you: Temporal landmarks spur goal initiation focus of attention. Psychological Review, 99(1), 172-83. when they signal new beginnings. Psychological Science, Pope, D., & Simonsohn, U. (2011). Round numbers as goals 0956797615605818. evidence from baseball, SAT takers, and the lab. Psychological Heath, C., Larrick, R.P., & Wu, G. (1999). Goals as reference science, 22(1), 71-79. points. Cognitive Psychology, 38, 79–109. Scott, M. L., & Nowlis, S. M. (2013). The effect of goal specificity Kivetz, R., Urminsky, O., & Zheng, Y. (2006). The goal-gradient on consumer goal reengagement. Journal of Consumer hypothesis resurrected: Purchase acceleration, illusionary Research, 40(3), 444-459. goal progress, and customer retention. Journal of Marketing Wright, P. M., & Kacmar, K. M. (1994). Goal specificity Research, 43, 39–58. as a determinant of goal commitment and goal Koo, M., & Fishbach, A. (2008). Dynamics of self-regulation: How change. Organizational Behavior and Human Decision (un)accomplished goal actions affect motivation.Journal of Processes, 59(2), 242-260. Personality and Social Psychology, 94, 183–195. Locke, E. A., & Latham, G. P. (1990), A Theory of Goal Setting and Task Performance. Englewood Cliffs, NJ: Prentice-Hall. Counterfeits Can Benefit Original Brands When People are Caught Using Counterfeits: The Role of Face Restoration Liangyan Wang, Shanghai Jiaotong University, USA Qin Wang, Arizona State University, USA L. Robin Keller, University of California Irvine, USA Eugene Chan, Monash University, Australia

EXTENDED ABSTRACT terms of the prestige that they convey, but they cost far less than origi- Counterfeit products are typically inferior-quality imitations of nals (Cordell et al. 1996; Grossman and Shapiro 1988). Purchasing genuine products, especially ones that carry a high brand value (Lai based on a desire to convey prestige to others means that, if they are and Zaichkowsky 1999), usually sold at a much lower price to be af- caught using counterfeits, consumers should feel embarrassed for us- fordable for everyday consumers. As for the impact of counterfeit on ing a product that only superficiallyappears prestigious but is actually the genuine products, previous research revealed mixed findings. On not. Embarrassment results from a public observation of an action the one hand, counterfeits can reduce consumers’ demand for genuine that others consider inappropriate; it threatens the positive public self- products, particularly luxury ones (Wilcox et al. 2009). However, Nia image that people convey; and it impairs the effectiveness of social and Zaichkowsky (2000) found that over 70% of survey respondents interactions (Dahl, Manchanda, and Argo 2001). The emphasis is on indicated that the availability of counterfeits does not decrease their public action, in that embarrassment is distinct from what an indi- faith in and purchase intention for genuine (luxury) brands. Thus, vidual feels alone (Keltner and Buswell 1997). Behaviorally, embar- there are different views regarding whether counterfeit products bene- rassment motivates individuals to take actions that restore their self- fit or harm the image and equity of genuine products. This incongruity image (Feinberg, Willer, and Keltner 2012), also known as restoring illustrates the complexity of understanding counterfeit consumption, face. According to Goffman (1967), face is a positive self-image that as well as the need for further study to understand its impact for both is affirmed through interaction with others. It is a social resource that consumers and marketers. is maintained, enhanced, or lost through interpersonal interactions. In the current investigation, we add to this literature by examin- In our case, being caught with counterfeits should result in embar- ing how a specific phenomenon, namely being caught with counter- rassment and drive consumers to restore their face because they are feits, would increase consumers’ preference for the genuine products. observed using products that go against socially-approved norms and We propose an interactive effect between consumers’ self-construal that elicit social disapproval. (interdependent vs. independent) and product type (symbolic vs. We hypothesize that one (but not the only) way to restore face functional). Specifically, we hypothesize that consumers with an – that is, engaging in facework (Jiang and Cova 2012) – is to opt interdependent self-construal who are caught using symbolic (vs. for the genuine version of the counterfeit product that consumers are functional) products increase their preference for genuine products caught using. Given that consumers caught using counterfeits are because they are embarrassed and wish to restore face – and one (but embarrassed because their products only convey prestige but are not not the only) way to do so is by consuming the genuine version of the actually prestigious, this should motivate them to restore their face by counterfeit product that they use or wear. We also predict no effect acquiring and possessing the product that is actually prestigious – that on preference for genuine products among consumers with an inde- is, the genuine article. pendent self-construal because they are not embarrassed when caught Interdependent Consumers and Symbolic Products using such products. We test and confirm our hypotheses in a series of However, not every consumer who is caught using counterfeit four experimental studies. products should demonstrate a greater preference for genuines, and she/he should not show a greater preference for genuines when caught Theoretical Background with any product. Rather, the preference for genuine products should Caught Using Counterfeit Products be strongest for consumers with interdependent (vs. independent) Research on counterfeit consumption has predominantly ex- self-construals who are caught using counterfeit symbolic (vs. func- plored its pre-purchase antecedents, such as how and why consum- tional) products. The distinction in product type is crucial (Penz and ers procure counterfeit products. For example, Wilcox et al. (2009) Stottinger 2005). Symbolic products allow consumers to express their argued that consumers buy counterfeits when the genuine product, actual or ideal self-image (Onkvisit and Shaw 1987). Functional prod- especially in the case of luxuries, provides a social-adjustive func- ucts provide consumers with utility or other functional benefits (Yoo, tion. Other studies have focused on other factors such as the product Chung, and Han 2006). It is important to note that the same product attributes and cultural norms that determine consumers’ decisions to can be either symbolic or functional depending on how consumers purchase counterfeit products (Eisend and Schuchert-Güler 2006). view and use it. A handbag can be symbolic since it offers feelings Interestingly, only a few studies have examined the post-purchase of status or professionalism, but it can also be functional because it phases. Nia and Zaichkowsky (2000) and Amaral and Loken (2016) allows women to carry everyday articles. Thus, given that symbolic investigated how the owners of genuine products feel and value their products are more important for consumers’ self-image, we firstly ex- authentic merchandise when they see that other consumers are sport- pect that consumers who are caught with counterfeit symbolic prod- ing counterfeit products. However, it is not known how consumers ucts should show a greater preference for genuine articles to than who buy, use, and/or wear counterfeit products themselves subse- those who are caught with counterfeit functional products. quently change their preference for genuine products. This motivated We secondly – and perhaps more importantly – predict that con- our present inquiry in which we focus specifically on consumers who sumers with interdependent self-construals should express a greater are caught using counterfeit products. preference for genuine articles when caught with a counterfeit sym- Consumers buy counterfeit products primarily because they typ- bolic product, compared to those with independent self-construals. ically resemble, superficially at least, genuine or luxury products in Purchase of symbolic products is strongest by consumers high on

Advances in Consumer Research 948 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 949 face consciousness (Chen et al. 2014), namely those with interdepen- action was significant (F = 3.89, p =.05). Subjects with interdependent self-construals (Hwang, Francesco, and Kessler 2003). Mem- dent self-construals indicated a more positive preference change for bers of interdependent cultures view the self as part of a collective a genuine Longines watch in the symbolic than functional conditions identity with others, and they view face as a socially-defined aspect. (Msymbolic = 1.17 vs. Mfunctional = .17, t = 2.87, p = .005). In contrast, stu- This concern for face makes Asian consumers less tolerant of social dents with independent self-construals did not show such a preference failures (Chan, Wan, and Sin 2009). Consumers with interdependent change for a genuine Longines watch (Msymbolic = .51 vs. Mutilitarian = .47, self-construals should be more embarrassed when caught using coun- p = .90). However, when the students were not caught by others, there terfeit symbolic products because they lose an aspect of the self that was neither a main effect nor a two-way interaction (ps > .55). they define and convey socially. They are likely to be seen as fru- Study 1 provided beginning evidence for our proposition. Since gal, vain, or that they intend to impress others without putting in the there is no effect in the condition where consumers use counterfeit but effort (Commuri 2009). This also makes it more likely that interde- are not caught by others, we only focused on the “caught” situation in pendent consumers who have been embarrassed when “caught in the the following studies. act” should express a stronger preference for genuine products, in comparison to independent consumers. Indeed, negative evaluations Study 2 motivate interdependent consumers to protect their public self-image Study 2 aims to replicate the results from Study 1and also dem- to maintain their well-being that is largely defined by their social self- onstrate that being caught with counterfeit symbolic products increas- worth (Dickerson, Gruenewald, and Kemeny 2004). es interdependent consumers’ embarrassment, which then mediates In sum, we hypothesize that interdependent (vs. independent) their preference for genuine products. consumers feel embarrassed when they are caught using counterfeit Method. This study used a 2 (product type: symbolic, functional) symbolic (vs. functional) products, motivating their preference for the × 2 (self-construal: interdependent, independent) between-partici- genuine versions as a means to restore face. We test our proposed pants design. 135 undergraduate students from a large university in framework in four studies. Australia participated in this study. We first primed self-construal using the Sumerian warrior story task (Trafimow, Triandis, and Goto Study 1 1991). The manipulation of product type was the same as in Study Study 1 demonstrated our main hypothesis that interdependent 1 except using a different brand Tissot. Then participants imagined (vs. independent) consumers caught with counterfeit symbolic (vs. being caught with a counterfeit Tissot watch at a friend’s party. Upon functional) products express a greater preference for genuine prod- completion, participants completed willingness to pay for genuine ucts, compared to a baseline condition in which they sport the coun- Tissot watch that we adapted from Rucker and Galinsky (2008). Par- terfeit products but are not caught using them. ticipants also indicated on separate 9-point scales (1 = “Not at all” to Method. This study used a 2 (self-construal: interdependent, 9 = Very much”) the extent they felt the following emotions at the independent) × 2 (product type: symbolic, functional) × 2 (caught: current instant: ashamed, embarrassed, insecure, and vulnerable. yes, no) between-participants design. Participants (n = 173) from a Results. As predicted, a 2 × 2 ANOVA on students’ willingness large university in China first finished a pronoun circling task which to pay (WTP) for a genuine Tissot watch revealed only an interaction primed them either an interdependent or independent self-construal between the two factors (F = 10.10, p < .002). Participants primed (Brewer and Gardner 1996). Then they randomly received one of two with interdependence expressed a higher WTP for genuine Tissot ad copy conditions for a Longines watch that highlighted either its watches when they imagined being caught with counterfeit symbolics symbolic or functional benefits, which we adapted from Wilcox et al. rather than with counterfeit functionals (Msymbolic = 6.26 vs. Mfunctional = (2009). Subsequently, participants were shown an actual counterfeit 4.30, t = 3.32, p < .001). No such effect emerged for participants with

Longines watch and asked to wear the watch, ostensibly as market independent self-construal (Msymbolic = 4.83 vs. Mfunctional = 5.64, p = research for Longines (the actual company) to understand consum- .21). A moderated mediation analysis with embarrassment as media- ers’ interest in counterfeit versions of its products. Afterwards, par- tor was significant (95% CI: -1.51, -.07), supporting our hypothesis. ticipants proceeded to a different room while wearing the counterfeit Longines watch, ostensibly to help another researcher with their Study 3 study while we were preparing the second part of the Longines watch In Study 3, we aim to demonstrate that being caught with coun- study. They were either “caught” or “not caught” by a confederate in terfeit symbolic products also increases their motivation to restore this unrelated task. We obtained participants’ preference change for a face, which then also mediates their preference for genuine articles. genuine Longines watch by subtracting the purchase intent measured Method. This study used a 2 (self-construal: interdependent, before the counterfeit Longines watch usage task from the intent mea- independent) × 2 (product type: symbolic, functional) between-par- sured afterwards. ticipants design, in which self-construal was measured and product Results. We first conducted manipulation check for our manipu- type was manipulated. Participants (n = 129) completed a similar lations of self-construal as well as product type and got results vali- manipulation of product type of iPhone 5s and the self-construal dating the manipulations. Then we conducted a 2 × 2 × 2 ANOVA and measurement (Singelis 1994). We also assessed participants’ motiva- found a main effect of being caught or not with a counterfeit watch, tion to restore face by asking them to indicate their agreement with with participants who were caught scoring higher on preference than the following statement on a single 7-point scale (1 = “Strongly dis- those not caught (β = 3.92, S.E. = 1.71, t = 2.29, p = .02). This of- agree” to7 = “Strongly agree”): “After being caught wearing a coun- fers evidence for our fundamental hypothesis that being caught with terfeit iPhone 5s, I can restore face by purchasing and using a genuine counterfeits in general can shift preference towards genuine products, iPhone”. relative to not being caught at all. Crucially, the three-way interaction Results. As predicted, the product type × self-construal interac- was marginally significant (F = 3.43, p = .07). As such, we analyzed tion was significant (β = .70,S.E. = .29, t = 2.43, p = .02). Specifically, the 2 (self-construal) × 2 (product type) interaction within each of the students with an interdependent self-construal (+1 S.D.) indicated a caught and not caught conditions to further decompose the effects. more positive preference for a genuine iPhone after getting caught When the participants were caught using counterfeits, the 2 × 2 inter- using counterfeit symbolics than counterfeit functionals (β = 1.04, 950 / Counterfeits Can Benefit Original Brands When People are Caught Using Counterfeits: The Role of Face Restoration

S.E. = .42, t = 2.78, p = .01). But for participants with an indepen- Dahl, Darren W., Rajesh V. Manchanda, and Jennifer J. Argo dent self-construal (-1 S.D.), there was no such difference p( = .31). A (2001), “Embarrassment in Consumer Purchase: The Roles moderated mediation analysis with face restoration as mediator was of Social Presence and Purchase Familiarity”, Journal of significant (95% CI: .03, .49). Consumer Research, 28 (3), 473-81. Dickerson, Sally S., Tara L. Gruenewald, and Margaret E. Kemeny Study 4 (2004), “When the Social Self is Threatened: Shame, In Study 4, we present a boundary condition for the effect. In- Physiology, and Health”, Journal of Personality, 72 (6), 1191- terdependent consumers no longer prefer genuine symbolic products 216. even when they are caught with them when they do not expect to see Eisend, Marin and Pakize Schuchert-Güler (2006), “Explaining again the other individuals who caught them. Counterfeit Purchases: A Review and Preview”, Academy of Method. The study used a 2 (expected re-encounter: yes, no) × 2 Marketing Science Review, 12(6), 1-25. (product type: symbolic, functional) × 2 (self-construal: interdepen- Feinberg, Matthew, Robb Willer, and Dacher Keltner (2012), dent, independent) between-participants design (n = 284). We ma- “Flustered and Faithful: Embarrassment as a Signal of nipulated self-construal and product type using the same tasks as in Prosociality”, Journal of Personality and Social Psychology, Study 2. In the scenarios in which participants imagined being caught 102 (1), 81-97. while wearing a counterfeited Tissot watch at a party, half of them Goffman, Erving (1967), Interaction Ritual: Essays on Face-to- specifically imagined that they expected to attend another party the Face Behavior, London: Doubleday. week after with the same party attendees, while the other half did not Grossman, Gene M. and Carl Shapiro. (1988). “Foreign have such an expectation. Counterfeiting of Status Goods”, Quarterly Journal of Results. A 2 × 2 × 2 ANOVA revealed that the three-way interac- Economics, 103 (1), 79-100. tion of product type, self-construal, and expected re-encounter was Hwang, Alvin, Anne Marie Francesco, and Eric Kessler (2003), significant (F = 3.73, p = .05). When participants expected to see the “The Relationship Between Individualism-Collectivism, same party attendees again, the 2 × 2 interaction was significant (F Face, and Feedback and Learning Processes in Hong Kong, = 5.44, p = .02). Participants primed with interdependence expressed Singapore, and the United States”, Journal of Cross-Cultural a greater preference for genuine Tissot watches when they imagined Psychology, 34 (1), 72-91. being caught with counterfeit symbolics rather than with counterfeit Jiang, Ling and Veronique Cova (2012), “Love for Luxury, functionals (Msymbolic = 5.88 vs. Mfunctional = 4.60, t = 2.43 p = .016). In Preference for Counterfeits: A Qualitative Study in Counterfeit contrast, participants primed with independence expressed a similar Luxury Consumption in China”, International Journal of preference whether they imagined being caught with counterfeit sym- Marketing Studies, 4 (6), 1-9. bolics or with counterfeit functionals (Msymbolic = 5.34 vs. Mfunctional = Keltner, Dacher and Brenda N. Buswell (1997), “Embarrassment: 5.46, p = .69). But when participants did not expect to see the same Its Distinct Form and Appeasement Functions”, Psychological party attendees again, there was no two-way interaction between self- Bulletin, 122 (3), 250-70. construal and product type (c). Lai, Kay Ka-Yuk and Judith Lynne Zaichkowsky (1999), “Brand In conclusion, this research focus on being caught with counter- Imitation: Do the Chinese Have Different Views?”, Asia feits, thus adding a component to this domain of study that is largely Pacific Journal of Management, 16(2), 179-92. missing in the literature. Our findings offer contributions for academ- Nia, Arghavan and Judith Lynne Zaichkowsky (2000), “Do ic scholars, policy makers, and marketing practitioners. Counterfeits Devalue the Ownership of Luxury Brands?”, Journal of Product & Brand Management, 9 (7), 485-97. REFERENCES Onkvisit, Sak and John Shaw (1987), “Self-Concept and Image Amaral, Nelson B. and Barbara Loken (2016), “Viewing Usage Congruence: Some Research and Managerial Implications”, of Counterfeit Luxury Goods: Social Identity and Social Journal of Consumer Marketing, 4 (1), 13-23. Hierarchy Effects on Dilution and Enhancement of Genuine Penz, Elfriede and Barbara Stottinger (2005), “Forget the ‘Real’ Luxury Brands”, Journal of Consumer Psychology, 26 (4), Thing – Take the Copy! An Explanatory Model for the 483-95. Volitional Purchase of Counterfeit Products”, Advances in Brewer, Marilynn B. and Wendi Gardner (1996), “Who is This Consumer Research, 32, 568-75. ‘We’? Levels of Collective Identity and Self Representations”, Rucker, Derek and Adam D. Galinsky (2008), “Desire to Acquire: Journal of Personality and Social Psychology, 71 (1), 83-93. Powerlessness and Compensatory Consumption”, Journal of Chan, Haksin, Lisa C. Wan, and Leo Y. M. Sin (2009), “The Consumer Research, 35 (2), 257-67. Contrasting Effects of Culture on Consumer Tolerance: Singelis, Theodore M. (1994), “The Measurement of Independent Interpersonal Face and Impersonal Fate”, Journal of Consumer and Interdependent Self-Construals”, Personality and Social Research, 36 (August), 292-304. Psychology Bulletin, 20 (5), 580-91. Chen, Yun-Qing, Hong Zhu, Meng Le, and Yi-Zhen Wu (2014), Trafimow, David, Harry C. Triandis, and Sharon G. Goto (1991), “The Effect of Face Consciousness on Consumption of “Some Tests of the Distinction Between the Private Self Counterfeit Luxury Goods”, Social Behavior and Personality, and the Collective Self”, Journal of Personality and Social 42 (6), 1007-14. Psychology, 60 (5), 649-55. Commuri, Suraj (2009), “The Impact of Counterfeiting on Genuine- Wilcox, Keith, Hyeong Min Kim, and Sankar Sen (2009), “Why Item Consumers’ Brand Relationships”, Journal of Marketing, Do Consumers Buy Counterfeit Luxury Brands?”, Journal of 73(3), 86-98. Marketing Research, 46 (2), 247-59. Cordell, Victor V., Nittaya Wongtada, and Robert L. Kieschnick Yoo, Shijin, Seh-Woong Chung, and Jin K. Han (2006), “A (1996), “Counterfeit Purchase Intentions: Role of Lawfulness Durable Replacement Model for Symbolic Versus Functional Attitudes and Product Traits as Determinants”, Journal of Consumption: An Integrated Cultural and Socio-Economic Business Research, 35 (1), 41-53. Perspective”, Global Economic Review, 35 (2), 193-206. Design an Experience Bundle: The Role of Experience Structure Juan Wang, University of Guelph, Canada Miranda Goode, Western University, Canada June Cotte, Western University, Canada

EXTENDED ABSTRACT ing structure would be more preferred was also found in this study. In Imagine planning a spring vacation and coming across an ad that addition, complementarity perception was found to mediate this pref- features a six-day voluntour trip. The ad suggests an itinerary where erence (Hayes 2013). Notably, the fact that the experience bundle was half of the trip involves volunteering in jungle restoration in the Ama- evaluated less favourably in the condition with descriptions on sepa- zon and the other half sightseeing in Peru. This is an example of an rate pages (vs. alternating) condition suggests that variety perception experience bundle, a bundle that is composed of two or more experi- is not the underlying mechanism. ences (e.g., volunteering and sightseeing) that are usually consumed In study 3, participants (N=175; between-subjects design) were independently of one another. To maximize desire for this bundle, an asked to list their thoughts about the sequencing of activities in an important decision must be made by marketers: how should the ex- educational trip to Singapore and then evaluated the experience. This periences in the bundle be structured? Should the trip be structured trip was a bundle of sightseeing and educational engagements, and such that consumers will complete sightseeing in the first three days was structured either alternately or sequentially. Participants’ thoughts before jungle restoration (i.e., sequential structure), or should the trip were coded for complementarity inferences. Greater number of com- be structured to include sightseeing experiences and jungle restoration plementarity inferences implies greater complementarity between (i.e., alternating structure) during each day? bundled experiences. As expected, complementarity perception medi- With three studies involving a variety of experience bundles, we ated preference for the educational trip, such that a greater number of find that alternately (vs. sequentially) structured experience bundles complementarity inferences were generated for the alternately struc- are more preferred. This is because an alternating structure allows con- tured trip. Further, some participants were provided with complemen- sumers to infer greater value with an experience bundle, which leads tarity inferences in the sequentially structured experience condition, to higher complementarity perception. For example, when the volun- and their evaluations increased in comparison to when complementar- tour trip is structured sequentially, consumers may infer that the jungle ity inferences were not provided. restoration experience will be spiritually rewarding and that the tour- To the best of our knowledge, we are the first to examine experi- ing experience will be fun. In other words, the multisensory elements, ence bundles, an increasingly popular market offering. Theoretically, events, and benefits of each experience are processed mainly inde- we contribute to the bundling literature in several ways. First, we show pendently of one another. In contrast, consumers may infer additional that complementarity perception, an important determinant in bundle benefits when the jungle restoration and sightseeing experiences are al- evaluation (Gaeth et al. 1990; Popkowski Leszczyc and Häubl 2010; ternated throughout the trip. Not only will consumers identify that the Stremersch and Tellis 2002), is quite malleable. Prior literature often jungle restoration will be spiritually rewarding and the sightseeing will assumes, implicitly, that complementarity perceptions are stable. We be fun, but they may also infer that boating along the Amazon River show these perceptions can be influenced by varying the structure of (touring event) will be relaxing after finishing a half-day of tree-plant- an experience bundle. More importantly, we show that complementar- ing (volunteer event). Further, appreciation of the river-maintenance ity perception is malleable because consumers engage in an inferential activity (volunteer event) on the following day may also be enhanced. process when learning about experience bundles. This process has not It is these inferred values that enhances complementarity perception received attention in the bundling literature. Equally important, we of- and high evaluations for alternately structured experience bundles. fer insights on how to better design and market experience bundles. In study 1, participants (N=110) evaluated a French Festival event in a between-subjects design. The festival was composed of three films REFERENCES (different genres) and three acrobatic shows (different themes), and Gaeth, Gary J., Irwin P. Levin, Goutam Chakraborty, and Aron was structured either alternately or sequentially. Supporting our expec- M. Levin (1990). “Consumer Evaluation of Multi-Product tations, the alternately (vs. sequentially) structured French festival was Bundles: An Information Integration Analysis,” Marketing perceived more favorably to consumers. Notably, a follow-up study Letters, 2(1), 47-57. with 80 participants found no differences in variety perceptions (Kahn Hayes, A.F. (2013). “An introduction to mediation, moderation, and and Wansink 2004) between the alternately and sequentially structured conditional process analysis: A regression-based approach,” bundles, ruling out this perception as an alternative explanation to our New York, NY: Guilford Press. hypothesized effect. Kahn, Barbara.E. and Brian Wansink (2004). “The Influence of In study 2, participants (N=97; between-subjects design) were Assortment Structure on Perceived Variety and Consumption asked to evaluate a full-day Toronto trip and indicate how comple- Quantities,” Journal of Consumer Research, 30(4), 519-533. mentary the trip seemed to be (Koukova, Kannan, and Kirmani 2012). Koukova, Nevena T., P.K. Kannan, and Amna Kirmani (2012). This trip was a bundle of two historic and two modern tours and was “Multiformat Digital Products: How Design Attributes Interact structured alternately or sequentially. We also included an inference- with Usage Situations to Determine Choice,” Journal of impairment condition, where subjects again evaluated an alternating Marketing Research, 49(1), 100-114. trip, but the individual tour descriptions were displayed on four sepa- Popkowski Leszczyc, Peter T.L. and Gerald Häubl (2010). “To rate pages. The descriptions, however, appeared on one page in the Bundle or not to Bundle: Determinants of the Profitability of alternating condition. In this condition, we expected that complemen- Multi-item Auctions,” Journal of Marketing, 74(4), 110-124. tarity perceptions would be reduced if participants had to flip back and Stremersch, Stefan and Gerard J. Tellis (2002). “Strategic Bundling forth between descriptions to identify meaningful connections among of Products and Prices: A New synthesis for Marketing,” the tours. Consequently, we expected the alternating trip would also be Journal of Marketing, 66(1), 55-72. perceived less favourably. Support for the expectation that an alternat-

EXTENDED ABSTRACT products (e.g., trucker hat, Mickey Mouse watch), and styles (e.g., Consumers typically covet the meanings associated with the punk, goth, grunge). The most common reason for consuming ironi- products they use (Belk 1988; Levy 1959). Rolex wearers want to cally was to be humorous, a motive mentioned by 45% of ironic con- seem classy, gangsta rap fans want to seem tough, and flag wavers sumers but none of the control consumers (χ2 = 41.89, p < .001. For want to seem patriotic. But sometimes consumers use a product ironic consumers to successfully be funny, however, the audience while trying to dissociate from the tastes, meanings, associations, or needs to both recognize that the consumer is being ironic and per- ideals that the product typically signifies. That is, they consume the ceive the ironic consumption to be humorous. Our subsequent stud- product ironically. For example, consumers wear “ugly Christmas ies thus investigate when observers detect ironic consumption and sweaters” to holiday parties, emulate infamous cultural figures (e.g., how they evaluate ironic consumers. Bill Cosby) on Halloween (DailyMail 2015), and watch TV shows like “The Jersey Shore” and “The Bachelor” that they consider cat- When is Ironic Consumption Detected? egorically awful (McCoy and Scarborough 2014; Thompson 2000). Under what conditions will people recognize that another con- Similarly, when Donald Trump announced his candidacy, liberal ur- sumer is being ironic? People detect verbal irony when they think banites made Trump’s red hat the “ironic accessory of the summer” that an utterance is the opposite of what the speaker intends (Utsumi (Parker 2015). 2000), just as they detect situational irony when an occurrence is Although there is an established literature on sincere symbol- the opposite of what they would normally expect (Lucariello 1994). ic consumption – that is, using products to signal who you are or We similarly expect that people will detect ironic consumption only avoiding products to signal who you are not (Belk, 1988; Berger and when the meaning of a consumption behavior is different than the Heath 2007; Douglas and Isherwood 1978; Veblen 1899) – far less meaning the consumer (hereafter, an incongruent product). However, is known about ironic consumption. We ask three questions in at- a product can be incongruent either because it has an aspirational in- tempt to bridge this gap in the literature: Why do people consume congruent meaning or dissociative incongruent meaning (Englis and ironically? When do observers detect ironic consumption? And, what Solomon 1995; White and Dahl 2006). Because consumers want to impression do ironic consumer makes on others? distance themselves from dissociative meanings but not aspirational meanings (Berger and Heath 2007; White and Dahl 2006), irony Why Consume Ironically? should be detected more frequently when a consumer uses a prod- The literature suggests several reasons why consumers might uct with a dissociative meaning. A popular high school student who use products ironically. Ironic consumption potentially signals status dresses like a nerd, for example, might seem ironic because others (Brooks 1981). Ironically adopting “kitch” products (e.g., reality TV, would recognize that (a) the popular kid is not a nerd, and (b) he soul food), might help distinguish knowledgeable consumers both probably doesn’t want to be a nerd. On the other hand, people should from lower class consumers, who genuinely like the products, and be less likely to detect irony when a nerd dresses like a popular kid less knowledgeable upper-class consumers, who don’t know about (an aspirational identity), a nerd dresses like other nerds (a congru- the product or how to consume it ironically (Bourdieu 1984; John- ent identity), or a popular kid dresses like other popular kids (also a ston and Baumann 2010). Another possibility is that irony offers con- congruent identity). sumers an indirect way to criticize social institutions, consumption practices, or people (Wompole 2012). Klein (2000), for example, Study 2 suggests that consumers can express dissent for commercialism by Study 2 tested the hypothesis that an observer is only likely to ironically consuming mass-marketed goods, like “Baywatch” or Dis- detect irony when a consumer uses an incongruent product with a neyworld (Klein 2000). A third possibility is that people consume dissociative meaning. Undergraduate students (N = 374) were ran- ironically to be humorous. People report using verbal irony to be domly assigned to a condition in a 2 (consumer: cool, uncool) x 2 funny (Roberts and Kreuz 1994); they may use products ironically (product: congruent, incongruent) x 2 (cool style: grunge, preppy) for the same reason. between-subjects experiment. Participants read about two high school students, “John” and Study 1 “Chris,” who are part of different social cliques. One clique was The primary purpose of study 1 was to examine consumers’ described as being cool upper classmen, the other as uncool fresh- motivations for consuming ironically. Participants from Mechanical man. The incongruent dissociative condition described a cool high Turk (N = 408) read a definition of ironic consumption, answered school student adopting a style associated with an uncool clique. The whether or not they had “ever used a product or brand to be ironic,” incongruent aspirational condition described an uncool high school and, if so, explained why they used the product ironically. As a con- student adopting a style associated with a cool clique. The congruent trol condition, half of the sample alternatively indicated if they had conditions described a cool student adopting a style associated with a “ever used a product or brand to be symbolic,” and, if so, explained cool clique and an uncool student adopting a style associated with an why. Two coders (89% agreement) indicated whether the responses uncool clique, respectively. To ensure that the effects did not depend mentioned one or more of the following motivations: status, criti- on the type of style adopted, we counterbalanced whether the cool cism, humor, identity signaling, belonging, signaling support, enjoy- clique was associated with a “grunge” style or a “preppy” style (the ment, functional benefits. results generalized across the two styles). Reports of ironic consumption were not rare, but only a mi- We measured irony detection in two ways. As a conservative nority of the sample ironically used a product (23%). Participants measure, we asked a general, open-ended question: “Based on what ironically consumed a range of brands (e.g., Hollister, Polly Pocket), you’ve read, why do you think John (Chris) chose to wear this to-

Advances in Consumer Research 952 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 953 day?” Two coders (100% agreement) assessed whether or not the product (i.e., the bumper sticker), we asked if they personally identi- participant mentioned irony or a synonym. We subsequently asked fied with the politician in the bumper sticker (either Bernie Sand- participants to indicate the extent to which they disagree (1) or agree ers or Donald Trump, depending on the condition) on a seven-point (7) with the following items: “He is being ironic,” and “He is wear- scale from “not a lot/1” and “a lot/7.” ing the clothing ironically.” Despite relatively low average levels of irony detection, both Irony detection was low overall, but it was most common when the open-ended and closed-ended measures revealed that irony was the consumer used an incongruent product linked to a dissociative most likely to be detected when the consumer used a product with group (i.e., when the cool kid dressed like uncool kids). In contrast, a dissociative meaning, in this case a sticker of a politician with op- participants detected less irony when the consumer dressed like an posing political views (ps < .01. Importantly, the effect of detecting aspirational incongruent group (i.e., the uncool kid dressing like cool ironic consumption on the impression of the consumer depended on kids) or an associative, congruent group (consumer-by-product inter- the participants’ identification with the political bumper sticker. A action: (F(1,366) = 12.16, p < .001. moderated mediation analysis (Hayes 2013; model 14) confirmed that the participant’s political identification moderated the mediat- What are the Consequences of Consuming Ironically? ing effect of irony detection on the extent to which the consumer Consumers use products ironically in attempt to be humorous, seemed both humorous (Index of moderated mediation = -.076; 95% but whether being ironic actually makes a consumer seem more or C.I.: -.13 to -.03) and likable (Index of moderated mediation = -.11; less humorous (as well as more or less likable) should depend on 95% C.I.: -.16 to -.06). Participants were significantly more likely to whether the audience identifies with the product being consumed. detect irony when the consumer used an incongruent rather than a Sincerely consuming a product signals that the consumer belongs to congruent product (b = .80; t = 7.94, p < .001). Moreover, the effect the same in-group as others who use the product (Berger and Heath of irony detection on the extent to which the consumer seemed both 2007; Douglas and Isherwood 1979). Irony, however, distances humorous (b = -.09; t = -5.01, p < .001) and likable (b = -.14; t = 7.86, the consumer from the product and, by extension, those who use p < .001) depended on the participant’s identification with the con- it sincerely. Because people have a more favorable impression sumed product. Participants who scored relatively low on identifica- of members of an in-group than members of an out-group (Tajfel tion with the consumed product rated the target consumer as more 1982), people who identify with a product should evaluate consum- humorous and more likable if they thought the consumer was being ers more favorably if they think the consumer is using the product ironic, whereas participants who scored relatively high on identifica- sincerely rather than ironically. For example, drivers with a Donald tion rated ironic consumers as being less likable (see figure 1). Thus, Trump bumper sticker should make a more favorable impression detecting irony blunted the positive (and negative) impressions that on Trump supporters than on people who do not support Trump, people typically have of consumers who use products with which but only if the consumption of the sticker seems sincere rather than they identify (do not identify). ironic. Study 4 Study 3 The purpose of study 4 was to examine the effects of ironic con- Study 3 attempted to replicate study 2 by showing that an audi- sumption on impressions by directly manipulating whether a con- ence is most likely to detect irony when a consumer uses a product sumer uses a product sincerely or ironically and whether or not the with a dissociative meaning. It also attempted to extend study 2 by audience identifies with the product being consumed. Participants examining the effects of being ironic on impressions from others. recruited from Mechanical Turk (N = 238) evaluated a consumer U.S. citizens on Mechanical Turk (N = 295) completed the who was wearing a tee-shirt either ironically or sincerely as part of study during a presidential primary race that featured democrat a 2 (consumption: ironic, sincere) x 2 (product identification: high, Bernie Sanders and republican Donald Trump. Leveraging the fact low) between-subjects experiment. that most Americans dissociate from people with opposing political Depending on the identification condition, participants either views (Murray et al. 2014), we created two dissociative incongru- listed a musician who they consider a favorite (high identification) ent conditions and two congruent conditions using a 2 (consumer: or who they dislike (low identification). Participants subsequently liberal, conservative) x 2 (product: liberal, conservative) between- imagined encountering a consumer wearing a tee-shirt of this musi- subjects design. cian and asking him or her about the shirt. In the “sincere consump- Participants viewed an image of an automobile with a politi- tion” condition, the consumer ostensibly responded: “I’m wearing cal bumper sticker and reported their impression of the driver. In the shirt because I’m a fan.” In the “ironic consumption” condition, the conservative-congruent condition the driver had a Donald Trump the person ostensibly responded: “I’m only wearing the shirt to be sticker on a Hummer with an NRA logo. In the conservative-incon- ironic.” Participants subsequently indicated how the interaction gruent condition, the Hummer instead displayed a Bernie Sanders would influence the extent to which they considered the consumer bumper sticker. In the liberal-congruent condition, the driver dis- humorous (e.g., less funny/more funny) and likable (e.g., less lik- played a Bernie Sanders sticker on a Toyota Prius with a peace sign. able/more likable). Finally, in the liberal-incongruent condition, the Prius displayed a Consistent with study 3, a 2 (consumption: ironic, sincere) x Donald Trump bumper sticker. A pretest confirmed that a Hummer, 2 (product identification: high, low) ANOVA revealed a significant the NRA, and Donald Trump were all associated with conservatives interaction on the extent to which the consumer seemed both humor- whereas a Prius, a peace symbol, and Bernie Sanders were all associ- ous (F(1, 234) = 13.52, p < .001) and likable (F(1,234) = 31.27, p ated with liberals. < .001). When participants identified with the musician on the tee- After viewing the car, participants completed the irony detec- shirt, they saw the consumer as being less humorous (F(1, 234) = tion measures from study 2 and indicated the extent to which they 12.22, p < .001) and less likable (F(1,234) = 23.03, p < .001) when considered the driver humorous (e.g., “He/She seems like a funny (s)he wore the tee-shirt ironically. In contrast, when participants did person) and likable (e.g., “I have a favorable impression of him/ not identify with the musician on the tee-shirt, they saw the consum- her”). Finally, to assess participants’ identification with the focal er as being marginally more humorous (F(1, 234) = 2.75, p < .10) 954 / Ironic Consumption and significantly more likable (F(1,234) = 9.41, p < .01) when (s)he Johnston, Josée, and Shyon Baumann, (2014), Foodies: Democracy wore the shirt ironically. Thus, wearing a shirt of a musician that the and Distinction in the Gourmet Foodscape, New York: audience likes (doesn’t like) led to a more (less) favorable impres- Routledge. sion of the consumer, but not if the consumer was being ironic. Klein, Naomi (2000), No Logo: taking Aim at the Brand Bullies, New York: Picador. General Discusson Levy, Sidney J. (1959), “Symbols for Sale,” Harvard Business Our research addresses a gap in the literature on symbolic con- Review, 37(4), 117-124. sumer behavior by investigating ironic consumption. Not all con- Lucariello, Joan (1994), “Situational Irony: A concept of events sumers engage in ironic consumption, but those who do say that they gone awry,” Journal of Experimental Psychology: General, use products ironically to be funny rather than to signal a particular 123(2), 129-145. identity or group membership. For example, a music snob might wear McCoy, Charles Allan, and Roscoe C. Scarborough (2014), a Justin Bieber tee-shirt ironically as a joke. Whether the audience “Watching “Bad” Television: Ironic Consumption, Camp, and recognizes and appreciates the joke depends on whether they detect Guilty Pleasures,” Poetics, 47, 41-59. the irony, and this is most likely to happen when they recognize that Murray, Alan, et al. (2014), “Political Polarization in the American the consumer wants to dissociate from the product’s conventional Public,” Pew Research Center Report, published June 12, meaning or identity. For example, people are more likely to think 2014, accessed 11/22/2016 from http://www.people-press. that the music snob is being ironic if they recognize that he disdains org/2014/06/12/political-polarization-in-the-american-public/ Justin Bieber and his “Beliebers” than if they think the music snob Parker, Ashley (2015), “Trump’s Campaign Hat Becomes an Ironic is or wants to be a Belieber. Importantly, rather than making a con- Summer Accessory,” New York Times, September 11, uploaded sumer seem funny outright, being ironic appears to blunt both posi- on 11/22/2016 from http://www.nytimes.com/2015/09/13/ tive and negative impressions from others. People who don’t care for fashion/trumps-campaign-hat-becomes-an-ironic-summer- Justin Bieber will think that Bieber shirt-wearers have a poor sense accessory.html. of humor, unless they think they are wearing the shirt ironically. On Roberts, Richard M., and Roger J. Kreuz (1994), “Why Do People the other hand, Beliebers will tend to like Bieber shirt-wearers more, Use Figurative Language?” Psychological Science, 5(3), 159- unless they think they are wearing the shirt ironically. By blunting 163. the impression that consumers make on others, ironic consumption Simmel, Georg (1957), “Fashion,” American Journal of Sociology, potentially offers a safer way to consume products with uncertain, 62(6), 541-558. contentious, or even undesirable meanings. Tajfel, Henri (1982), “Social Psychology of Intergroup Relations,” Annual Review of Psychology, 33(1), 1-39. REFERENCES Thompson, Craig J. (2000), “Postmodern Consumer Goals Made Easy,” The Why of Consumption: Contemporary Perspectives Belk, Russell W. (1988), “Possessions and the Extended Self,” on Consumer Motives, Goals, and Desires, New York: Journal of Consumer Research, 15(2), 139-168. Routledge, 120-39. Berger, Jonah, and Chip Heath (2007), “Where Consumers Diverge Utsumi, Akira (2000), “Verbal Irony as Implicit Display of from Others: Identity Signaling and Product Domains,” Ironic Environment: Distinguishing Ironic Utterances from Journal of Consumer Research, 34(2), 121-134. Nonirony.” Journal of Pragmatics 32(12), 1777-1806. Bourdieu, Pierre (1984), Distinction: A Social Critique of the Veblen, Thorstein (1899), The Theory of the Leisure Class, New Judgement of Taste, Harvard University Press. York: The New American Library. Brooks, John (1981), Showing Off in America, Boston/Toronto: Wampole, C. (2012), “How to Live Without Irony,” New York Little, Brown and Company Limited. Times, November 18, SR1. Douglas, Mary, and Baron Isherwood (1978), The World of Goods: White, Katherine, and Darren W. Dahl (2006), “To Be or Not Be? Towards an Anthropology of Goods, London: Allen Lane. The Influence of Dissociative Reference Groups on Consumer Englis, Basil G., and Michael R. Solomon (1995), “To Be and Preferences,” Journal of Consumer Psychology, 16(4), 404- Not to Be: Lifestyle Imagery, Reference Groups, and the 414. Clustering of America,” Journal of Advertising, 24(1), 13-28. Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach, New York, NY, US: Guilford Press. The Biographies of Things: Provenance and the Pursuit of Moral Consumption Todd Weaver, Point University, USA Pam Ellen, Georgia State University, USA Carolyn Curasi, Georgia State University, USA

EXTENDED ABSTRACT resources, particularly the natural resources contained in products or Anecdotal evidence suggests that consumers are increasingly those used or affected by the provenance of products. One example concerned with the background of the products they consume. These is David’s concern about the natural resources used in building his product biographies relate to the provenance of products: the people, own housing community. David notes the irony of building or pur- places, resources, and processes involved in creating and distributing chasing something new to be more sustainable: the new thing uses products for consumption. In this research, we show that some con- up more of the Earth’s scarce resources, which is often contradictory sumers want to know the biographies of products they buy so they to sustainability. can choose items with backgrounds consistent with their moral val- The third type of provenance biography mentioned by our in- ues, such as environmental concerns or social justice. Furthermore, formants, person biographies, concerns the people involved in prov- we find product biographies can create tensions and dilemmas for enance. Our informants frequently expressed concern for the work- consumers as they attempt to balance multiple consumption goals ers involved in the production of consumer goods. For example, with their moral values. Dorothy is uncertain about who specifically benefits from Fair Trade Kopytoff (1986) introduced the idea of a product biogra- certification and how exactly they might benefit, but she still seeks phy as a way of conceptualizing meanings relating to a product’s Fair Trade certification with the hope that she might be helping the background, proposing that products, like people, have “life histo- workers in the supply chain in terms of work conditions and pay. ries” that inform and influence the meanings attached to them by Meanings associated with the geographical background, place consumers. He notes that product biographies would have many biographies, constituted a fourth type of provenance biography facets covering a wide range of culturally-constituted and objective mentioned by our informants. In some instances, informants even meanings, including where a product is made and by whom, and that discussed distant or abstract places they had never visited. Lacey, a biographical meanings can have important moral implications for working mother with young children, is uncertain exactly where the consumers. Our goal in this research was to explore whether con- apple juice she buys at Publix comes from, but infers that the United sumers sought biographical information when purchasing and, if so, States is “a pretty good place” for making apple juice and that Indo- to understand the types of biographical information that consumers nesia is less desirable. seek. We also sought to understand how these biographies impact Our informants noted that product biographies could pose a consumption decisions as consumers attempt to consume morally. number of dilemmas in their decision making. Informants frequently To explore product biographies and their impact on consum- encountered a conflict between moral values, such as environmental- ers, we selected the pseudonymous Springs Landing as the site for ism and altruism, and pragmatic goals to conserve time and money. our research, a neighborhood within a major metropolitan area in For example, Lacey feels great moral conflict when she shops at the southeastern U.S., where residents agree to live by certain com- Walmart: munity values, such as “Diversity” and “Sustainability.” Twenty-five residents were interviewed, and follow-up interviews were con- Our location makes it hard because there is a Walmart super- ducted with 10 of them. The sample included diverse age groups, store right around the corner. And with a child, sometimes it’s household types, and professions. Semi-structured (i.e., open-ended just easier to go to the Walmart superstore, and that place is just and non-directive) interviews were used to allow flexibility to ex- full of evilness. (Lacey, 30s, healthcare professional) plore novel or unexpected themes. Individual interviews were coded Like most consumers, Lacey values convenience and will some- to illuminate themes within each individual discourse. After coding times prioritize it over competing objectives. In this case, her beliefs individual transcripts, a thorough analysis of the data set as a whole about the biographies of the products available at Walmart make her was conducted. The overall thematic structure emerged via consen- feel guilty about shopping there because she feels she is not being sus of the research team. As a test of the credibility of the analysis, true to her moral consumption identity. we performed member checks by meeting with informants after re- Another conundrum was presented by prices. Several infor- search team consensus was reached (Hirschman, 1986; Lincoln & mants noted that the low prices of some consumer goods did not Guba, 1986). make sense given their provenance biographies. For example, Bill, Informants repeatedly referred to various processes involved a retiree living on a fixed income, reported that he primarily shops in systems of provenance in accounts of their consumption. Process at a local organic grocer, in part, because he feels that prices at na- biographies stemmed from all stages of provenance, from the ex- tional grocery chain Publix and other mainstream stores do not fairly traction of raw materials to the retailing of finished goods. Respon- represent the value of the natural resources required to create the dents frequently used these meanings to distinguish the morality of products. For Bill, shopping for groceries brings to mind the pro- one product from another. For example, the PlayStation video game duction processes involved, a thought that compels him to shop for system brought to mind negative meanings related to the materials groceries that are more expensive, despite his fixed income and their necessary for their construction for Neil. Although it did not prevent less convenient location. him from buying a game system, he clearly feels guilt for consuming something whose provenance might involve human suffering, even REFERENCES though he is not confident in this knowledge. Hirschman, Elizabeth C. (1986), “Humanistic Inquiry in Marketing Informants also demonstrated an awareness of product biog- Research: Philosophy, Method, and Criteria,” Journal of raphies related to the resources invested in making and marketing Marketing Research, 23 (August), 237-249. products. These resource biographies frequently related to natural

Kopytoff, Igor (1986), “The Cultural Biography of Things: Lincoln, Yvonna S., and Egon G. Guba (1986), “But is it Commoditization as Process,” The Social Life of Things: Rigorous? Trustworthiness and Authenticity in Naturalistic Commodities in Cultural Perspective, 68, 70-73. Evaluation,” New Directions for Program Evaluation, 30, 73-84. Possession Substitutability: Identity and Usage of Rented (Versus Owned) Products Liad Weiss, University of Wisconsin-Madison, USA

EXTENDED ABSTRACT To rule-out a “mere-involvement” account, whereby, when Despite availability of good rental options, many consumers identity is central, consumers care more about what gear they use still prefer to haul or ship their own gear for various activities (e.g., and prefer their own because they think it is better, study 2 manipu- skiing). Understanding psychological antecedents for consumer pref- lated the rental’s quality. Further, to promote a causality argument, erence for owned (versus rented) gear is pivotal for marketers in the identity-centrality was manipulated. Participants imagined owning access-based industry. However, such antecedents remain understud- high-quality fishing gear they used when fishing was central to their ied. identity. Subjects in the [“identity-central” / “identity-peripheral”] Previous research on consumer preference for owned (vs. un- condition read that fishing is [still / no-longer] central to their iden- owned) items finds that, when a good is more identity-relevant, own- tity. Then, after finding damage to their gear before an upcoming ers respond more negatively to its loss (Dommer and Swaminathan fishing trip, participants chose between paying to expedite repairs 2013; Ferraro, Escalas, and Bettman 2011). Relatedly, we propose and renting similar gear (DV). Subjects in the [higher / lower] qual- that when identity (e.g., skier) is more self-central, a consumer will ity condition learned that the rental gear was [better / worse] than the perceive relevant possessions (e.g., skis) as less substitutable due to gear they owned. Consistent with low substitutability, fishing iden- the classification of such possessions as more self-central (i.e., more tity centrality (vs. peripherality) (1) lowered renting preference and “me;” Weiss and Johar 2013). Consequently, the consumer will be (2) attenuated the positive effect of higher (vs. lower) quality of the less inclined to use rented gear to temporarily substitute a possession rental on renting likelihood. Mediation analysis showed that identity even when the rental is better. Thus, the very consumers who could centrality increased gear classification as “me,” and subsequent pos- most appreciate using superior gear (e.g., “die-hard” skiers) would session unsubstitutability – expectation that the experience would be be ironically the least likely to do so (due to being the most reluctant different without one’s possession. Consequently, subjects expressed to forgo using their own gear). Indeed, such ironic consequences of lower desire to rent regardless of the rental’s quality. identity directionally align with the aforementioned prior research. Studies 4a and 4b tested the moderating role of whether the ac- However, that prior research studied the preference for owned goods tivated identity is product-centric, or whether it pertains to activities only in the context of a possible loss of ownership over the good that require gear (e.g., listening to music as an “audiophile” requires (e.g., due to sale; Kahneman, Knetsch, and Thaler 1990). In con- hi-fi speakers), or product-agnostic, or whether it pertains to activi- trast, we uniquely test preference for owned goods in choices that ties that can be engaged in directly through the senses (e.g., listening eliminate a possible loss of ownership, specifically in dilemmas of to music as a “musicophile” only requires the ability to hear). Acti- whether to temporarily substitute an owned item. vating a product-agnostic identity was predicted to hinder construal Consistent with low possession substitutability as the driver for of objects in terms of “self” (Weiss and Johar 2016). Consequently, low renting, we predicted that impeding the effect of identity central- consumers were predicted to be unmindful of object “me-ness,” and ity on low possession substitutability will reverse the ironic effect of thus not to perceive possessions with higher “me-ness” as less sub- identity. We theorized that construing the self in terms of an identity stitutable. To test this prediction, Study 4a participants in the [“mu- that does not depend on gear for engaging in identity activities (e.g., sicophile” / “audiophile”] condition wrote two things about them- a “chef” requires equipment to cook, but a “foodie” only needs her selves that reflect their [music-lover / audiophile] side. Participants taste buds to eat) will impede the effect of identity centrality on pos- then imagined owning high quality headphones. In the scenario, session substitutability. In such situations, greater identity centrality they were online, buying tickets to a silent concert, “where music may actually predict more (vs. less) usage of superior rentals. is transmitted via radio waves and the audience listen to it through Results from five studies were consistent with our predictions. headphones.” Subjects chose between paying to use a high-fidelity In study 1, while planning an out-of-state cycling vacation, partici- radio connector they could plug their own headphones to (option 1) pants learned that the local shop rents out bikes identical to theirs. and paying a bit more to use the event’s wireless headphones, known The rental cost was either higher or lower than the cost of shipping for their better sound reproduction (option 2, DV). Results showed their bike there. After choosing between shipping and renting (DV), that greater “audiophile” centrality predicted lower renting when the participants reported the extent to which cycling was central to their activated identity was “audiophile,” but higher renting when the ac- identity in real-life. Consistent with lower possession substitutability, tivated identity was “musicophile.” Study 4b replicated the results identity centrality predicted (1) lower renting and (2) smaller posi- in incentive-compatible settings. Consequences for identity research tive effect of lower (vs. higher) renting price on renting likelihood. and for marketers in the booming access-based economy are dis- To rule-out a “renting-aversion” account, whereby identity cen- cussed. trality simply increases dislike of the unfamiliar, study 2 manipulated whether the gear in one’s possession (i.e., the familiar gear) was REFERENCES owned or borrowed. Further, to rule out a “sunk-cost” explanation Dommer, Sara L. and Vanitha Swaminathan (2013), “Explaining the involving initial investment in gear, any initial investment in the gear Endowment Effect through Ownership: The Role of Identity, was eliminated. Participants in the [“borrowed” / “owned”] condi- Gender, and Self-Threat,” Journal of Consumer Research, tion imagined planning a cycling vacation, deliberating between Forthcoming. shipping there the quality bike family friends recently left for them Ferraro, Rosellina., Jennifer E. Escalas, and James R. Bettman [to use for the year they are away / as a hand-me-down right before (2011), “Our Possessions, Our Selves: Domains of Self- going away for a year] (option 1) and renting gear at the local shop Worth and the Possession-Self Link,” Journal of Consumer (option 2). Consistent with predictions, identity centrality predicted Psychology, 21 (2), 169-77. lower renting when the gear in one’s possession was owned, but not when it was borrowed.

Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler (1990), --- (2016), “Products as Self-Evaluation Standards: When Owned “Experimental Tests of the Endowment Effect and the Coase and Unowned Products Have Opposite Effects on Self- Theorem,” Journal of Political Economy, 98 (6), 1325-48. Judgment,” Journal of Consumer Research, Forthcoming. Weiss, Liad and Gita V. Johar (2013), “Egocentric Categorization and Product Judgment: Seeing Your Traits in What You Own (and Their Opposite in What You Don’t),” Journal of Consumer Research, 40 (1), 185-201. The Influence of Purchase Motivation on Assortment Size Preferences Sarah C. Whitley, Boston University, USA Remi Trudel, Boston University, USA Didem Kurt, Boston University, USA

EXTENDED ABSTRACT ined purchasing a CD to listen to for a class assignment while those Consumers’ decisions to purchase a product are often driven in the hedonic condition considered purchasing a CD to listen to for by different motivations. Personal pleasures (i.e., hedonic - motiva pleasure. 56% of participants in the hedonic motivation condition tions) trigger consumption in some cases and functional needs (i.e., preferred to choose from a large (versus small) assortment, whereas utilitarian motivations) drive consumption in others (Botti and Mc- only 24% of participants in the utilitarian purchase motivation condi- Gill 2011; Choi and Fishbach 2011) ― even when the product is the tion preferred the large assortment (χ2(1) = 9.17, p < .01). Study 1B same. In the present research, we ask a novel question: do consumers uses a similar class-based manipulation, design, and assortment size prefer different assortment sizes when choosing a product for a he- measure for the purchase of a book. 65% of participants purchasing donic versus a utilitarian motivated purchase? the book to read for pleasure (hedonic) preferred the large assort- We argue that the number of alternatives consumers plan to re- ment, whereas only 27% of participants purchasing the book for a view for a purchase is influenced by whether their motivation for that class assignment (utilitarian) felt similarly (χ2(1) = 18.89, p < .01). purchase is hedonic versus utilitarian. More specifically, we believe Studies 1C and 1D replicate this main effect of purchase motivation that consumers exhibit different perceptions about the uniqueness of on assortment size using computers and paint colors, respectively. their ideal product preferences under hedonic versus utilitarian pur- In study 2, participants imagined selecting a song to listen to as chase motivations, which in turn affect the number of products they part of an online study (utilitarian) or as break between studies (he- review. We argue that consumers believe that their product prefer- donic). Participants the number of songs they would like to choose ences for hedonic purchases are inherently more unique and differ- from (between 2-24), before being presented with a selection of ent from those of other consumers, due to the idiosyncratic nature songs matching that size. Participants then selected a song, listened of hedonic purchase preferences. Consequently, consumers believe to it, and rated their satisfaction with the song and the song assort- they will have a more difficult time finding a product matching their ment. In addition to replicating our main effect of hedonic versus preferences under a hedonic purchase motivation, leading them to utilitarian motivation on assortment size (MH = 8.76, SD = 7.03 vs. consider a larger assortment of product alternatives. In contrast, con- MU = 5.43, SD = 5.62 ; F(1, 149) = 10.31, p < .01) using real choice, sumers with utilitarian purchase motivations tend to perceive their there were no differences in satisfaction with the song or the assort- product preferences as less unique compared to others and thus an- ment. This null effect on satisfaction provides confidence that our ticipate less difficulty in preference matching, resulting in an inclina- pre-choice assortment size scenarios accurately reflect consumers’ tion to choose from smaller assortments. real preferences. Across 7 studies, we find that consumers with hedonic purchase Lastly, we argue that consumers with hedonic purchase motiva- motivations prefer larger assortments than those with utilitarian pur- tions tend to perceive their preferences as unique (compared to oth- chase motivations. Our results show that consumers exhibit differ- ers’ preferences) to a greater extent than consumers with utilitarian ences in their perceptions of preference uniqueness and the level of purchase motivations. As a result, consumers with hedonic versus anticipated difficulty in preference matching between the hedonic utilitarian purchase motivations anticipate greater difficulty in find- and utilitarian purchase motivations, and these two variables under- ing a product that matches their preferences, resulting in predilection lie the effect of purchase motivation on preferred assortment size. for larger assortments. In study 3A, we measure both perceived pref- Our research contributes to the literature in three important erence uniqueness and anticipated difficulty in preference matching ways. We extend the literature on hedonic and utilitarian purchases to evaluate our sequential process model (purchase motivation  by uncovering how these two purchase motivations influence con- perceived preference uniqueness  anticipated difficulty in prefer- sumers’ preference for assortment size. Second, we add to the lit- ence matching  assortment size). Participants were randomly as- erature on the determinants of consumers’ preferences for large and signed to a hedonic utilitarian or motivation for the book purchase, small assortments. Lastly, a major contribution of our research is our analogous to study 1B. Participants then indicated their perceptions finding that hedonic and utilitarian purchase motivations have dif- of preference uniqueness as agreement (7-point Likert) with the fol- ferent implications for consumers’ perceptions of uniqueness about lowing statement: “I believe my preferences for books are unique and their ideal product preferences. different from others’ book preferences.” Participants subsequently Study 1A serves as an initial test for the effect of purchase mo- indicated how difficult they felt it would be to find a book that satis- tivation on assortment size preference. Similar to Botti and McGill fies their goal (1=Not at all Difficult, 7=Very Difficult). Lastly,- par (2011), we manipulate the purchase motivations for the same prod- ticipants stated how many books they would like to review in an uct, but emphasize the hedonic or utilitarian nature of the end-goal of assortment. The effect of purchase motivation on assortment size was the purchase. Since hedonic and utilitarian purchases are motivated replicated (F(1, 208) = 4.96, p < .05). Additionally, participants in the by a desire to experience affective pleasure and by a desire to ful- hedonic condition felt their preferences were more unique (F(1, 208) fill a need or accomplish a task (Dhar and Wertenbroch 2000; Pham = 10.32, p < .01; MH = 4.58, SD = 1.28 vs. MU = 3.96, SD = 1.51) and 1998; Strahilevitz and Myers 1998), respectively, we characterize perceived finding a preference matching product to be more difficult hedonic and utilitarian purchase motivations in our scenarios as plea- (F(1, 208) = 5.95, p < .05; MH = 4.07, SD = 1.73 vs. MU = 3.50, SD = sure-seeking and task-driven. Participants were assigned to either a 1.66) than participants in the utilitarian condition. Bootstrapping re- utilitarian or hedonic purchase motivation condition and asked their sults from a serial multiple mediator analysis (Model 6; Hayes 2013) preference between a small (6) or large (24) assortment of product confirmed a positive and significant indirect effect (β = .13; 95% CI alternatives. Participants were told to imagine they taking a class on = (.05, .35)) verifying our full mediation pathway. We replicated our Classical music; those in the utilitarian motivation condition imag- process model in study 3B using the same paint color manipulation

Advances in Consumer Research 959 Volume 45, ©2017 960 / The Influence of Purchase Motivation on Assortment Size Preferences as study 1D. Alternative explanations of maximizing/satisficing, Hayes, Andrew F. (2013), Methodology in the Social Sciences: anticipated enjoyment of choosing, efficiency of choosing, intrinsic Introduction to Mediation, Moderation, and Conditional motivation, and goal importance were explored but not supported. Process Analysis: A Regression-Based Approach, New York, NY: Guilford Press. REFERENCES Pham, Michel T. (1998), “Representativeness, Relevance, and the Botti, Simona and Ann L. McGill (2011), “The Locus of Choice: use of Feelings in Decision Making,” Journal of Consumer Personal Causality and Satisfaction with Hedonic and Research, 25, 144-59. Utilitarian Decisions,” Journal of Consumer Research, 37, Strahilevitz, Michal and John G Myers (1998), “Donations to 1065-78. Charity as Purchase Incentives: How Well they Work may Choi, Jinhee and Ayelet Fishbach (2011), “Choice as an End Versus Depend on what You are Trying to Sell,” Journal of Consumer a Means,” Journal of Marketing Research, 48, 544-54. Research, 24 (March), 434-46. Dhar, Ravi and Klaus Wertenbroch (2000), “Consumer Choice between Hedonic and Utilitarian Goods,” Journal of Marketing Research, 37 (Feb.), 60-71. Gifts of Consolation: Gifts as Substitutes for Emotional Support Hillary Wiener, SUNY-Albany, USA Tanya Chartrand, Duke University, USA

EXTENDED ABSTRACT Study 4 examines whether the substitution effect works in re- Most research on gifts has examined celebratory gifts, or gifts verse: If a giver gives a gift, are they then less likely to provide the that are given in response to others’ positive events, such as birthdays recipient with emotional support? Participants (n = 203) first read or holidays. However, consumers also give gifts in response to oth- that a friend’s (coworker’s) mother had died and that they decided ers’ negative life events, such as a loved one’s death or a break-up. to send flowers and a sympathy note or just a note. Then they were In this research, we introduce the term “gifts of consolation” to refer asked to write the note they would send. These notes were coded for to gifts given in response to negative events and examine when and how emotionally supportive they were. Participants who gave flow- why consumers give them. ers wrote significantly less supportive notes than did those who did We propose that gifts of consolation are a form of social support not give flowers (p < .0001). This effect was moderated by a marginal and thus that giving a gift can be a substitute for providing some- interaction (p = .09) such that giving a gift led to less supportive one with emotional support. People have the existing psychological notes in the friend condition (p < .0001) but only marginally less ability to substitute a product for a psychological construct (Chen, supportive ones in the coworker condition (p = .07). This shows that Wan, and Levy 2017), and we argue that giving gifts and providing givers compensate for not giving a gift by providing more emotional emotional support both satisfy a person’s obligation to provide social support. support to close others (Clark 1987). Therefore, if givers are unable Study 5 looks at when givers prefer to give gifts rather than pro- to provide emotional support, they should be more likely to give a vide emotional support. We propose that when a giver wants to sup- gift and vice versa. port someone who they think will be difficult to support (e.g., some- Studies 1a and 1b test the basic effect, examining whether giv- one who is negative or has low self-esteem; Forest, Kille, Wood, and ers spend more on gifts when have not provided emotional support. Holmes 2014), they will give a gift to fulfill their obligation to provide support while avoiding an unpleasant interaction. We test this Participants (nStudy1a = 208; nStudy1b = 209) were asked to imagine that their friend’s mother died and that they were able or unable to pro- idea in a lab study. Participants (n = 48) were told that they would vide them with emotional support. Then in study 1a, participants provide another participant with social support and were shown the chose between giving their friend one of three arrangements of sym- profile of the person they would support. The profile was actually pathy flowers (small—$40, medium—$60, and large—$80) or not not from another participant but manipulated how easy the person sending flowers, and in study 1b they chose how much they wanted was to support by varying how pessimistic and low in self-esteem to donate to a memorial fund for their friend’s mother. Participants they were. Then participants were asked if they wanted to spend five who provided less emotional support condition spent more on gifts, minutes talking to the person about their break-up (i.e., provide emotional support) or pay $1 to give the person a cookie and avoid the than did those who provided more support (pStudy1a = .006; pStudy1b = .029), suggesting that givers use gifts to compensate for not provid- conversation (i.e., give a gift). Participants were significantly more ing emotional support. likely to provide support by giving a gift when the support recipient Prior research (Clark 1987) has found that people feel less ob- sounded hard to support (54%) than easy to support (25%, p = .039), ligated to provide social support to people they are distant from, so showing that gifts of consolation are an appealing way to support in study 2 we examine whether the givers’ closeness to the support difficult people. recipient moderates the effect. Participants (n = 414) first read that In conclusion, we show that givers give gifts of consolation and a friend’s (close condition) or coworker’s (distant condition) mother treat them as substitutes for emotional support. Our findings suggest had died, and they attended (did not attend) the funeral. Then they that thinking about gifts as a form of social support may be a fruitful were asked how much they would donate to a memorial fund. There direction for future research. was an interaction between the level of emotional support provided and closeness (p = .06). Participants in the friend condition donated REFERENCES significantly less money ($54) when they attended the funeral than Chen, Rocky Peng, Echo Wen Wan, and Eric Levy (2017), when they did not (p < .0001), but the effect was marginal in the co- “The Effects of Social Exclusion on Consumer Preference worker condition ($21; p = .1). When feelings of obligation are low, for Anthropomorphized Brands,” Journal of Consumer givers do not compensate for a lack of emotional support with a gift. Psychology, 27 (1), 23-34. Study 3 tests whether givers substitute gifts for emotional sup- Clark, Candace (1987), “Sympathy Biography and Sympathy port in celebratory situations or whether this effect is unique to gifts Margin,” American Journal of Sociology, 93 (2), 290-321. of consolation. Participants (n = 415) read about a friend’s positive Forest, Amanda, David R. Kille, Joanne V. Wood, and John G. event (another friend was throwing them a birthday party) or nega- Holmes (2014), “Discount and Disengage: How Chronic tive event (their mother died), and that they did or did not attend the Negative Expressivity Undermines Partner Responsiveness birthday party (funeral). Then they indicated how much they would to Negative Disclosures,” Journal of Personality and Social spend on a gift for their friend. There was a significant interaction be- Psychology, 107 (6), 1013-1032. tween gift giving occasion and level of emotional support provided (p < .0001) such that participants said they would spend more on a gift when they did not attend their friend’s mother’s funeral than when they did (p < .0001), but whether or not they had attended their friend’s birthday party did not influence how much they chose to spend on a gift (p = .36), showing that givers substitute more for negative than positive events.

Advances in Consumer Research 961 Volume 45, ©2017 You are not as Smart as You Think You are: Effects of Self-Perceived Knowledge on Consumer Information Processing and Decision Making Ruomeng Wu, University of Cincinnati, USA Brianna Escoe, University of Cincinnati, USA Frank R. Kardes, University of Cincinnati, USA

EXTENDED ABSTRACT identify a lack of information. However, when consumers are high It is well established that product knowledge affects informa- in both actual and self-perceived knowledge, the impact of self- per- tion processing and decision-making. For this reason, it has received ceived knowledge is greater because of our general tendency to ig- extensive attention from marketers (Bettman and Park 1980; Brucks nore missing information. Finally, self-perceived knowledge is sus- 1985; Johnson and Russo 1984; Moorman, Diehl, Brinberg, and ceptible to change through manipulation. Kidwell 2004). To test these hypotheses, five studies were conducted. In Study There are two distinguishable types of knowledge that affect 1, we examined the impact of self-perceived knowledge on choice how consumers process knowledge and make decisions: actual behavior. 136 participants viewed an advertisement of a bottle of knowledge (i.e. objective knowledge) and consumers’ assessments wine. We found that given that the advertisement was favorable, of their knowledge (i.e. subjective knowledge) (Bearden, Hardesty, regardless of objective knowledge, as self-perceived knowledge in- and Rose 2001; Moorman, Diehl, Brinberg, and Kidwell 2004). The creased, participants chose to purchase the wine more often (p < .05; two types of knowledge uniquely influence information search and 1 = Yes, 2 = No). choice behavior (Radecki and Jaccard 1995). In Study 2, we investigated the mediating role of omission ne- In this article, we focus on self-perceived knowledge. Specifi- glect in attitude extremity and purchase intention. 102 participants cally, we investigate how self-perceived knowledge affects purchas- were invited to evaluate a bottle of wine based on an advertisement. ing behavior and decision-making through omission neglect, igno- Regression analysis of self-perceived knowledge showed that par- rance of missing information. ticipants with higher self-perceived knowledge had more extreme Detecting absent information is surprisingly difficult (Sanbon- attitudes (p < .01), had higher purchase intentions (p < .0001), and matsu, Kardes, Houghton, Ho, and Posavac 2003; Sanbonmatsu, perceived the information as more sufficient (p = .001). Mediation Kardes, Posavac, and Houghton 1997; Sanbonmatsu,Kardes, and analysis (Hayes, 2012; Model 4; Bootstrap: 5000) showed that per- Sansone 1991). Consumers often rely heavily on the given infor- ceived information sufficiency mediated the path from self-perceived mation and ignore the possibility of missing information or under knowledge to attitude extremity (95%; CI: .0275 to .1395) and to weigh its importance even when they are aware it is missing. The purchase intention ((95%; CI: .0720 to .3040). failure to detect the absence of relevant information can encourage In Studies 3a and 3b, we changed participants’ vulnerabil- consumers to form more extreme judgments because the importance ity to omission neglect by manipulating participants’ self-perceived of the presented information is overestimated (Sanbonmatsu et al. knowledge. Study 3a worked as the preliminary test for Study 3b. In 2003; Unkelbach, Fiedler, and Freytag 2007). Overlooking impor- Study 3a, 83 participants viewed information about a camera, and tant omissions often leads to extreme judgments and poor decisions, we found that participants who reported that they knew more about which hinder the consumer. These decisions are often biased to the cameras perceived the information as more sufficient (p < .0001) and presented information. Because omission neglect is consequential, it held more extreme evaluations (p < .0001). is crucial to discover its antecedents. However, limited research has In Study 3b self-perceived knowledge was manipulated rather investigated the antecedents of omission neglect. than measured. 153 participants were randomly divided into either a Prior research has demonstrated that objective knowledge re- high or low self- perceived knowledge condition. Participants in the duces omission neglect (Hernandez, Han, and Kardes 2014). In con- low self-perceived knowledge condition were asked if they were pro- trary, we propose that self-perceived knowledge increases attitude fessional photographers. We expected most participants would report extremity and purchase likelihood by enhancing consumers’ suscep- that they were not professional photographers. Next, participants tibility to omission neglect. For instance, when evaluating a bottle of reported how much they knew about cameras on a scale biased to wine, if consumers are confident in their knowledge of wine, they are induce them to believe that they had a limited amount of knowledge more likely to focus on given information at the expense of absent about cameras (1 = very little, 5 = some). information. They believe that the given information is sufficient, Participants in the high self-perceived knowledge condition form a highly favorable judgment, and decide to purchase the wine were asked if they had taken pictures using a camera. We expected based solely off the positive description given in an advertisement. most participants to have taken pictures using a camera. Participants This research sheds light on the following questions: Does self- then reported how much they knew about cameras on a scale biased perceived knowledge affect omission neglect differently from and to induce them to believe that they had high knowledge of cameras more strongly than objective knowledge? How do these differences (1 = some, 5 = very much). shape judgments and decisions? Can we manipulate self-perceived Participants in the high (vs. low) self-perceived knowledge con- knowledge in order to change how consumers process information? dition perceived the information as more sufficient (p < .05) and held We argue that consumers who perceive themselves as highly more extreme evaluations (p = .05). knowledgeable form more extreme attitudes and are more likely to In Study 4, a response latency task testing memory for previ- purchase the product based on favorable product stimuli. This hap- ously presented (vs. absent) attribute information of a camera was pens because self-perceived knowledge increases consumers’ vul- used to show the direct effects of self-perceived knowledge on omis- nerability to omission neglect. Whereas consumers high in objective sion neglect. Out of 35 participants, those with greater self-perceived knowledge process information more cautiously and are more apt knowledge were more vulnerable to omission neglect. We see this to identify a lack of information, consumers high in self-perceived because they recognized previously present attributes much faster knowledge process information less cautiously and are less apt to than absent attributes (p < .05).

In this research, we offer a new, theoretical account of how self- Moorman, Christine, Kristen Diehl, David Brinberg, Blair perceived knowledge affects consumer information processing and Kidwell (2004). “Subjective Knowledge, Search decision-making through omission neglect. Locations, and Consumer Choice,” Journal of Consumer Research, 31(December), 673-680. REFERENCE Radecki, Carmen M. and James Jaccard (1995). “Perceptions Bearden, William O., David M. Hardesty, and Randall L. Rose of knowledge, actual knowledge, and information search (2001). “Consumer Self‐Confidence: Refinements in behavior,” Journal of Experimental Social Psychology, Conceptualization and Measurement,” Journal of Consumer 31(March), 107-138. Research, 28(June), 121-134. Sanbonmatsu, David M., Frank R. Kardes, David C. Houghton, Bettman, James R. and Whan C. Park (1980). “Effects of Prior Edward A. Ho, and Steven S. Posavac (2003). “Overestimating Knowledge and Experience and Phase of the Choice Process the importance of the given information in multiattribute on Consumer Decision Processes: A Protocol Analysis,” consumer judgment,” Journal of Consumer Psychology, Journal of Consumer Research, 7(December), 234-248. 13(March), 289-300. Brucks, Merrie (1985). “The Effects of Product Class Knowledge Sanbonmatsu, David M., Frank R. Kardes, Steven S. Posavac, on Information Search Behavior,” Journal of Consumer and David C. Houghton (1997). “Contextual Influences on Research, 12(June), 1-16. Judgment Based on Limited Information,” Organizational Hayes, Andrew F. (2012), “PROCESS: A Versatile Computational Behavior and Human Decision Processes, 69(March), 251- Tool for Observed Variable Mediation, Moderation, and 264. Conditional Process Modeling,” White Paper, http:// www. Sanbonmatsu, David M., Frank R. Kardes, and Carol Sansone afhayes.com/introduction-to-mediation-moderationand- (1991). “Remembering less and inferring more: Effects of time conditional-process-analysis.html#process. of judgment on inferences about unknown attributes,” Journal Hernandez, Jose Mauro C., Xiaoqi Han, and Frank R. Kardes of Personality and Social Psychology, 61(October), 546-554. (2014). “Effects of the perceived diagnosticity of presented Unkelbach, Christian, Klaus Fiedler, and Peter Freytag (2007). attribute and brand name information on sensitivity to missing “Information repetition in evaluative judgments: Easy to information,” Journal of Business Research, 67(May), 874- monitor, hard to control,” Organizational Behavior and 881. Human Decision Processes, 103(May), 37-52. Johnson, Eric J. and Edwards J. Russo (1984). “Product Familiarity and Learning New Information,” Journal of Consumer Research, 11(June), 542-550. Feel Closer When Event Comes: Time-moving Metaphor and Temporal Estimation Xiaobing Xu, Nankai University, China Rong Chen, Tsinghua University, China Miaolei (Liam) Jia, National University of Singapore, Singapore

EXTENDED ABSTRACT nario which was similar to the one used in Study 1 with two excep- In the domain of time, various metaphorical expressions of tions. First, the module representation was replaced by a presentation time are borrowed from space. Two spatiotemporal metaphors attract for a job interview. Second, no specific attitude towards the interview particular attention: ego-moving metaphor in which we are actively was indicated. We manipulated event time by setting the job inter- moving towards a fixed target event along the timeline; event-mov- view either on “next Monday” or “last Monday.” The time-moving ing metaphor in which the target event is moving towards us who are metaphor manipulation was modified from Boroditsky (2000), and fixed (Boroditsky 2000; Matlock et al. 2011; Richmond, Wilson, and we presented participants with pictures showing a human figure Zinken 2012). and a box. After that, participants reported their perceived temporal The current research explores a novel consequence of time- distance as in Study 1. Results showed that for future events, par- moving metaphors: the perceived temporal distance from now to the ticipants perceived the event to be farther away in the ego-moving event. For example, suppose you are going to the dreamy Maldives condition than in the event-moving condition. However, for the past next month for vacation. Should you feel psychologically closer or event, there was no significant difference in perceived temporal dis- farther away to this event when the departure day is constructed as tance between the ego-moving condition and the event-moving con- event-moving (e.g., the departure day is coming) versus ego-moving dition. (e.g., you are going to the departure day)? Study 3 tested the two accounts of time-moving metaphor on Specifically, we propose that the event-moving metaphor will emotional arousal. MTurk participants (N=143) were randomly as- make consumers perceive a higher level of emotional arousal than signed to three conditions (ego-moving, event-moving, and control the ego-moving metaphor. Our reasoning is twofold. First, people condition). Similar to study 1, all participants were first asked to read exert the motion action in the ego-moving metaphor, whereas the a scenario which was about preparing a birthday party for their best event exert the action in the event-moving metaphor. Therefore, friend. We used the same manipulation of time-moving metaphors people have a higher level of agency beliefs in the ego-moving meta- as in Study 1. In the control condition, the event was described as phor, which reduces people’s emotional arousal. Second, the moving stationary (e.g., “With the party being one week away, I hope…”, state of the event in the event-moving metaphor may induce a higher and “The party is in a few days; I think I should…”). Temporal dis- level of threat perception. The event is moving toward people in the tance and emotional arousal were measured in the same way as in event-moving metaphor, while it is static in the ego-moving meta- study 1. Results revealed that participants perceived the event to be phor. From an evolutionary perspective, humans are more alert and farther away in the ego-moving condition than both in the event- stimulated to moving objects than to still ones. Therefore, people in moving condition and in the control condition. There was no signifi- the event-moving metaphor perceive a higher threat, which might cant difference in temporal distance between the event-moving and increase people’s emotional arousal. Furthermore, based on the idea the control conditions. These results indicated that the difference in that emotional arousal decreases perceived temporal distance (Van emotional arousal in the two time-moving metaphors was caused by Boven et al. 2010), we propose that people in the event-moving met- agency beliefs, rather than by perceived threat. Mediation analysis aphor may perceive the temporal distance from now to the event to confirmed the mediational role of emotional arousal in the effect of be closer than those in the ego-moving metaphor. time-moving metaphor on perceived temporal distance. The results of three studies provide converging support for our To the best of our knowledge, this research provides the first proposition. In study 1, undergraduate student participants (N=57) evidence that time-moving metaphor can influence consumers’ tem- were randomly assigned to the two conditions. All participants were poral estimation of an event. The current research also complements first asked to read a scenario which described a presentation for an existing research on time-moving metaphor and emotion. Prior re- important module next Monday. They were told that they did not search has demonstrated that the two time-moving metaphors induce like the interview. Time-moving metaphors were primed by asking emotions with different valences: ego-moving metaphor induces hap- participants to read two different descriptions of the day (e.g., “As I piness while time-moving metaphor induces anxiety and depression approach the day of the oral presentation, I hope……” in the ego- (Richmond et al. 2012). Our research finds that time-moving meta- moving condition; “As the day of the oral presentation approaches, phors also affect the other dimension of emotion, namely arousal. I hope……” in the event-moving condition). We used two items to measure participants’ perceived temporal distance to the presenta- REFERENCES tion: (1) “How far away does the presentation feel to you?” (1 = Anderson, Craig A., Kathryn B. Anderson, and William E. Deuser right away, 10 = far away), and (2) “How temporally distant do you (1996), “Examining an Affective Aggression Framework: feel the presentation is?” (1 = very close, 10 = very distant). To mea- Weapon and Temperature Effects on Aggressive Thoughts, sure participants’ emotional arousal, we adapted the scale used in Affect, and Attitudes,”Personality and Social Psychology Anderson et al. (1996). Results revealed that participants perceived Bulletin, 22(4), 366–376. the event to be farther away in the ego-moving condition than in the Boroditsky, Lera (2000), “Metaphoric Structuring: Understanding event-moving condition. A mediation analysis revealed that emotion- Time through Spatial Metaphors,” Cognition, 75(1), 1–28. al arousal mediated the effect of time-moving metaphor on temporal Matlock, Teenie, Kevin J. Holmes, Mahesh Srinivasan, and distance perception. Michael Ramscar (2011), “Even Abstract Motion Influences Study 2 employed a 2 (event time: future vs. past) × 2 (time- the Understanding of Time,” Metaphor and Symbol, 26(4), moving metaphor: ego-moving vs. event-moving) between-subjects 260–271. design. MTurk participants (N=134) were first asked to read a sce-

Richmond, Jill, J. Clare Wilson, and Jörg Zinken (2012), “A Van Boven, Leaf, Joanne Kane, A. Peter McGraw, and Jeannette Feeling for the Future: How Does Agency in Time Metaphors Dale (2010), “Feeling Close: Emotional Intensity Reduces Relate to Feelings?” European Journal of Social Psychology, Perceived Psychological Distance,” Journal of Personality and 42(7), 813–823. Social Psychology, 98(6), 872–885. “That’s bitter!” Culture Specific Effects of Gustatory Experience on Judgments of Fairness and Advancement Jialiang Xu, University of Manitoba, Canada Fang Wan, University of Manitoba, Canada Norbert Schwarz, University of Southern California, USA Yuhuang Zheng, Tsinghua University, China

EXTENDED ABSTRACT effort in a challenging tasks context and measures of fairness in a job A rapidly growing body of research shows that incidental senso- rejection context. As expected, Chinese participants who tasted bitter ry experiences can influence consumers’ thoughts, feelings, and be- lotus seeds reported higher motivation and more effort in response to haviors. These influences can reflect a number of different processes, challenging tasks than those who tasted candy or water (simple ef- including the use of sensory experiences as a source of information; fect on Chinese for motivation F [2,159] =7.201, p=.001; for efforts the influence of sensory processes on mental simulations; and the F [2,159] =4.756, p=.01). These effects were not observed among grounding of abstract concepts in metaphorically related concrete Canadian participants (simple effect on Canadians for motivation F experiences (for reviews, see Barsalou, 2008; Krishna & Schwarz, [2,159] =.641, p=.528; for efforts F [2,159] =.386, p=.680; interac- 2014; Landau, Meier, and Keefer 2010). In this work, we focus on a tion effect for motivation F [2,159] = 5.253, p=.006; for efforts F consumption related sensory experience that has received little atten- [2,159] = 3.722, p=.026). Also as expected, Canadian participants tion, namely bitter taste. judged the described job situation as more unfair after they tasted An analysis of metaphors and sayings related to bitter taste in- something bitter than sweet or neutral (simple effect on Canadians dicates large cultural differences. Most notably, in the Chinese lan- F [2,159] = 2.927, p=.05). The fairness judgments of Chinese par- guage and culture, bitterness is associated metaphorically with con- ticipants were not significantly influenced (simple effect on Chinese cepts of adversity. Chi-Ku (“eating bitterness”) means “to endure a F [2,159] =.354, p=.702) but showed a similar pattern (interaction hardship”, “to overcome difficulties”, and “to forge ahead” (Loyalka, effect F [2,159] =.702, p=.488). This may reflect that the Chinese stu- 2012). In China, young children are taught to embrace the experi- dents had acquired some of the cultural associations between injus- ence of Chi-Ku because it is regarded as a necessary step to progress tice and bitterness during their studies in Canada. Study 3 addressed towards their goals and achieve success, as illustrated by sayings this possibility. such as, “only via eating bitterness can you surpass your peers.” This Study 3. 72 native speakers of English from a major Canadian metaphorical meaning of bitter tastes is uncommon in the English university and 88 native speakers of Chinese from a top Chinese uni- language and unfamiliar to Canadians. Instead, Canadians associ- versity were exposed to one of three taste experiences (bitter, sweet, ate bitterness with unfair treatment and social injustice. This link is neutral) and then responded to the job rejection paragraph used in exemplified by a famous quote from Emmeline Pankhurst, an early Study 2. Replicating Study 2, Canadians found the job situation less British feminist leader, who observed, “I have not personally suf- fair after tasting something bitter than after tasting something sweet fered from the deprivations, the bitterness and sorrow which bring so or neutral (simple effect on Canadians F [2, 154] =4.505, p = .013). many men and women to a realisation of social injustice.” These ob- Chinese participants were not influenced by the taste manipulation servations suggest an association between bitter taste and endurance (simple effect on Chinese F [2, 154] = 1.283, p = .283), resulting of hardship in Chinese (but not in Canadian) culture and between in the predicted interaction between culture and taste, F(2, 154) = bitter taste and injustice in Canadian (but not in Chinese) culture. 3.249, p = .041. Therefore, we conducted three study to test the culture-specific In combination, these findings highlight that the same sensory metaphoric effects of bitter taste. We hypothesized that bitter taste (1) experience can have divergent culture-specific effects on judgment. influences Chinese (but not Canadian) participants’ judgments of en- Consistent with culture-specific metaphors, incidental bitter tastes durance and (2) Canadian (but not Chinese) participants’ judgments influenced Canadian, but not Chinese, assessments of social justice of fairness. and fairness and Chinese, but not Canadian, assessments of moti- Study 1. To assess taste associations within each culture, we vation and effort in a difficult situation. A comparison of Chinese asked 57 native Canadian students to rate how strongly sweetness, students in China and Canada further suggests that culture-specific bitterness, sourness, and saltiness are associated with injustice and metaphors can be acquired through prolonged cultural exposure. 87 native Chinese students to rate how strongly these tastes are as- How culture specific metaphors moderate the influence of identical sociated with hardship and challenges (Chan, Tong, Tan, and Koh, sensory experiences, and how such metaphors are acquired through 2013). As expected, Canadians associated injustice more with bitter- cultural exposure, provides a promising avenue for future research in ness (M=5.05) than sweetness (M=1.63), sourness (M=4.02) or salti- embodiment and sensory marketing. ness (M=3.67; all p<.001); Chinese associated hardship more with bitterness (M=5.77) than sweetness (M=3.28) sourness (M=4.09) or REFERENCES saltiness (M=3.29; all p<0.001). Barsalou, L. W. (2008). Grounded cognition. Annu. Rev. Psychol., Study 2. To test the expected culture-specific behavioral effects 59, 617-645. of bitter taste we conducted a 3 (taste: bitter/water/sweet) x 2 (cul- Chan, K. Q., Tong, E. M., Tan, D. H., & Koh, A. H. (2013). What ture: Canadian/Chinese) x 3 (judgment: fairness/effort/motivation) do love and jealousy taste like?. Emotion, 13(6), 1142-1149 factorial experiment, with the last factor manipulated within partici- Krishna, A., & Schwarz, N. (2014). Sensory marketing, pants. Participants (N = 165; average age 20, 55.5% female) were embodiment, and grounded cognition: Implications for students at a major Canadian university; 84 were native speakers of consumer behavior. Journal of Consumer Psychology, 24, Chinese and 81 were native speakers of English. They were first ex- 159-168. posed to one of three randomly assigned taste experiences (bitter, sweet, neutral) and then responded to measures of motivation and

Landau, M. J., Meier, B. P., & Keefer, L. A. (2010). A metaphor- Loyalka, M. (2012). Eating Bitterness: Stories from the Front Lines enriched social cognition. Psychological bulletin, 136(6), of China’s Great Urban Migration. Univ of California Press. 1045. 968 / Appreciation of Creativity: Unintended Consequences of Thematic versus Taxonomic Product Organization Appreciation of Creativity: Unintended Consequences of Thematic versus Taxonomic Product Organization Alison Jing Xu, University of Minnesota, USA Joan Meyers-Levy, University of Minnesota, USA Barbara Loken, University of Minnesota, USA Ryan Tzushuo Wang, University of Minnesota, USA

EXTENDED ABSTRACT ing, characterized by analytic context-independent thinking whereby Retailers can organize their products either taxonomically individual items are set contrastively against others. or thematically. Taxonomic product assortment involves organiz- Upon integrating these observations, we propose the following. ing broad classes of products hierarchically into increasingly more Because interdependents (but not independents) are chronically in- specific categories, where members in each product category share clined and possibly more adept at processing data relationally, the many features and frequently serve as substitutes for one another. For impact of thematic product assortments on a consumer’s relational example, the category of socks may consist of various low cut, ankle, processing mind-set and its downstream effect on appreciation of and knee-high socks. In contrast, products are thematically organized creative goods may be more likely to occur when individuals possess if they perform complementary roles in a consumption situation (Es- an interdependent than an independent self-construal. tes et al. 2012). To illustrate, although socks, an MP3 player, and We conducted three experiments where in each, we manipulated deodorant share few, if any, common features, they are thematically whether individuals examined several thematic or taxonomic prod- related because they are commonly used together during workouts. uct assortments. Subsequently we investigated the downstream effect These two types of product assortments are perceived by con- this exerted on individuals’ appreciation of creative ads or products sumers differently (Diehl, van Herpen, and Lamberton 2015). Taxo- encountered later in an unrelated context. Experiment 1 found that nomic assortments help consumers efficiently locate and compare participants who previously examined several thematic versus taxo- substitute products that interest them, potentially enhancing their sat- nomic product assortments exhibited both greater appreciation of isfaction with the purchase process and the goods they buy, whereas creative ads and they better understood them. thematic assortments increase exposure to complementary products Experiment 2 extended on this by examining the anticipated that may otherwise be overlooked by consumers. Studies that have moderating effect of self-construal, and instead of examining cre- compared the two types of assortment have found that thematic prod- ative ads, study 2 investigated the downstream effect of the alterna- uct assortments are perceived as more attractive, but also more ef- tive product assortments on individuals’ willingness-to-pay (WTP) fortful to process (Diehl, et al., 2014). Still, to date, little is known for both creative and conventional products. Consistent with ex- about how taxonomic and thematic product assortments influence the pectations, treatment effects emerged only on individuals’ WTP for processing of consumers who examine them and any downstream ef- creative products, not conventional ones. Further, individuals with fects such processing may have on their responses to goods observed an interdependent self-construal reported higher WTP for creative in unrelated contexts. products when they had previously examined the thematic, not the We reasoned that the related versus compartmentalized nature taxonomic product assortments. However, individuals with an inde- of products displayed in thematic versus taxonomic product assort- pendent self-construal reported equivalent WTP sums regardless of ments may induce consumers who examine them to adopt a rela- the products assortments they had examined earlier. tional versus an item-specific processing mind-set. Moreover, extant Finally, experiment 3 investigated whether a mediated-mod- research has established that each of these types of mind-sets can eration effect of relational processing (mediator) and self-construal persist and carry-over, potentially producing downstream effects on (moderator) can explain how product assortment influences consum- consumers’ perceptions and/or behavior when they later encounter ers’ appreciation of creative products. The procedure and treatments products in unrelated contexts (Kim and Meyers-Levy 2008). Perti- used in this study were the same as those of experiment 2 except nent to this, substantial evidence suggests that relational processing, that only creative products were included, WTP was replaced by which entails encoding connections among often distally related con- product evaluations, and an indicator was added to gauge individu- cepts, boosts not only creativity but appreciation of it (Runco 1991; als’ use of relational processing by asking them to solve 10 Remote Yang et al. 2011). Thus, we propose that consumers who examine Associates Test (RAT, Mednick 1962) problems. Because success- thematic (versus taxonomic) product assortments will elicit a rela- ful performance on RAT problems requires engaging iteratively in tional processing mindset that produces downstream enhancement relational processing (Kray, Galinsky, and Wong 2006) participants’ in these consumers’ appreciation of innovative, novel, or otherwise relational processing was gauged by their number of correctly solved creative goods, which typically carry higher profit margins. RAT problems. As anticipated, results revealed parallel interactive Nevertheless, certain individual differences may modify effects of product assortment and self-construal on both -individu whether such downstream effects occur in later contexts. Research als’ evaluations of the creative products and on the indicator of their on self-construal has distinguished individuals with an interdepen- relational processing. In addition, bootstrap analysis indicated that dent orientation, who define themselves more holistically in terms among interdependent individuals, the effect of product assortment of their relationships with others, from individuals possessing an in- was mediated by relational processing. However, this was not the dependent orientation, who define themselves more analytically by case among independent individuals. focusing on their internal attributes such as traits or abilities (Markus In sum, this research contributes to the literature on both mind- and Kitayama 1991). Further and importantly, Zhu and Meyers-Levy sets and product assortments by demonstrating that examination of (2009) established that these alternative ways of thinking about the thematic (but not taxonomic) product assortments can trigger a rela- self more generally shape cognition, such that interdependents seem tional processing mindset, and particularly among individuals pos- to gravitate toward relational processing that emphasizes connec- sessing an interdependent self-construal, this mindset carries over tions and holism, whereas independents favor item-specific process- and boosts appreciation of creative products and ads encountered later in unrelated contexts. The current research also has important Advances in Consumer Research (Volume 45) / 969 practical implications. It suggests that retailers can use thematic Kray, Laura J, Adam D Galinsky, and Elaine M Wong (2006). product assortments to increase consumers’ appreciation and pos- “Thinking within the Box: The Relational Processing Style sible purchase of frequently highly profitable innovative, novel, or Elicited by Counterfactual Mind-Sets.” Journal of Personality otherwise creative products that may be featured elsewhere in the and Social Psychology 91 (1): 33–48. doi:10.1037/0022- retail venue. 3514.91.1.33. Markus, Hazel R., and Shinobu Kitayama (1991). “Culture and the REFERENCES Self: Implications for Cognition, Emotion, and Motivation.” Bowden, Edward M, and Mark Jung-Beeman (2003). “Normative Psychological Review 98 (2): 224–53. doi:10.1037/0033- Data for 144 Compound Remote Associate Problems.” 295X.98.2.224. Behavior ResearchMethods, Instruments,& Computers 35 (4): Meyers-Levy, Joan, and Rui (Juliet) Zhu (2007). “The Influence 634–39. doi:10.3758/BF03195543. of Ceiling Height: The Effect of Priming on the Type of Diehl, Kristin, Erica van Herpen, and Cait Lamberton (2015). Processing That People Use.” Journal of Consumer Research “Organizing Products with Complements versus Substitutes: 34 (2): 174–86. doi:10.1086/519146. Effects on Store Preferences as a Function of Effort and Monga, A B, and Z Gurhan-Canli (2012). “The Influence of Mating Assortment Perceptions.” Journal of Retailing 91 (1). New Mind-Sets on Brand Extension Evaluation.” Journal of York University: 1–18. doi:10.1016/j.jretai.2014.10.003. Marketing Research 49 (4): 581–93. doi:10.1509/jmr.09.0515. Estes, Zachary, Michael Gibbert, Duncan Guest, and David Runco, M. A (1991). Divergent thinking. Ablex Publishing. Mazursky (2012). “A Dual-Process Model of Brand Extension: Triandis, Harry C., and Michele J. Gelfand. (1998). “Converging Taxonomic Feature-Based and Thematic Relation-Based Measurement of Horizontal and Vertical Individualism and Similarity Independently Drive Brand Extension Evaluation.” Collectivism.” Journal of Personality and Social Psychology Journal of Consumer Psychology 22 (1). Society for Consumer 74 (1): 118–28. doi:10.1037/0022-3514.74.1.118. Psychology: 86–101. doi:10.1016/j.jcps.2011.11.002. Winterich, Karen Page, and Michael J Barone (2011). “Warm Estes, Zachary, Sabrina Golonka, and Lara L. Jones (2011). Glow or Cold, Hard Cash? Social Identify Effects on Thematic Thinking. The Apprehension and Consequences of Consumer Choice for Donation Versus Discount Promotions.” Thematic Relations. Psychology of Learning and Motivation - Journal of Marketing Research 48 (5): 855–68. doi:10.1509/ Advances in Research and Theory. Vol. 54. doi:10.1016/B978- jmkr.48.5.855. 0-12-385527-5.00008-5. Yang, Xiaojing, Torsten Ringberg, Huifang Mao, and Laura a. Friedman, R S, and J Forster (2001). “The Effects of Promotion and Peracchio (2011). “The Construal (In)compatibility Effect: Prevention Cues on Creativity.” Journal of Personality and The Moderating Role of a Creative Mind-Set.” Journal of Social Psychology. doi:10.1037/0022-3514.81.6.1001. Consumer Research 38 (4): 681–96. doi:10.1086/660118. Golonka, Sabrina, and Zachary Estes (2009). “Thematic Zhao, Xinshu, John G. Lynch, and Qimei Chen (2010). Relations Affect Similarity via Commonalities.”Journal of “Reconsidering Baron and Kenny: Myths and Truths about Experimental Psychology. Learning, Memory, and Cognition Mediation Analysis.” Journal of Consumer Research 37 (2): 35 (6): 1454–64. doi:10.1037/a0017397. 197–206. doi:10.1086/651257. Hunt, R. Reed, and Gilles O. Einstein (1981). “Relational and Item- Zhu, Rui, and Joan Meyers-Levy (2009). “The Influence of Self- Specific Information in Memory.” Journal of Verbal Learning View on Context Effects: How Display Fixtures Can Affect and Verbal Behavior 20 (5): 497–514. doi:10.1016/S0022- Product Evaluations.” Journal of Marketing Research 46 (1): 5371(81)90138-9. 37–45. doi:10.1509/jmkr.46.1.37. Kim, Kyeongheui, and Joan Meyers-Levy (2008). “Context Effects in Diverse-Category Brand Environments: The Influence of Target Product Positioning and Consumers’ Processing Mind‐Set.” Journal of Consumer Research 34 (6): 882–96. doi:10.1086/524309. Effects of Age on Taste Perceptions in a Food Service Context Amanda Pruski Yamim, Grenoble École de Management, France JoAndrea Hoegg, University of British Columbia, Canada Darren Dahl, University of British Columbia, Canada Adilson Borges, NEOMA Business School, France

EXTENDED ABSTRACT tresses and by reinforcing their age manipulation by informing con- The population is aging. In 1990 9.2 percent of the population sumers that Tiane is 25 (55) years old before the tasting. Study two was 60 or older; by 2013 it was 11.7, and by 2050 it is expected provided evidence that food prepared by an older (versus younger) to reach 21.1 (United Nations 2013). The aging of population has person is perceived to taste better and leads to higher purchase inten- significant implications for the workforce. The 55+ segment is tion. Further, the effect of age on taste perception was mediated by growing three times the rate of U.S. labor force, and is expected to participants’ perceived preparation time, whereby a slower prepara- reach 25% of the labor force by 2024 (Toossi 2015). At first glance tion time led participants to believe that the food tasted better. it would seem that this predicted workforce expansion could pro- In study 3 (N=217 workers from an European panel) we clarify vide an employment opportunity for older workers and vice versa. the direction of the age effect (if it was the fact that the chef was Alarming, elderly people face important challenges in the workplace older that increased taste perception or if was the fact that she was (Kaufmann, Krings, and Sczesny 2015). They are perceived to be younger that decreased taste perceptions) by adding a control group inefficient, slower to learn, and poorer at remembering than younger where the age of the cook was not informed. This study results show counterparts, (DeArmond et al. 2006; Ng and Feldman 2012). In the that the age effect seems to be driven by the perception that food pre- food service industry, where speed and efficiency are valued, percep- pared by an older person is better, rather than the perception that food tions that older employees are slower or get confused can result in a prepared by a younger person is worse. Further, study 3 provided belief that older workers can decrease customer satisfaction (Bae and evidence that the effect of age on taste is caused by the assumption Dae-Young 2014). In other words, restaurant managers tend to ad- that the food took longer to be prepared. here to the negative stereotype that old equals slow, and assume that Finally, in study four we identified a boundary condition for slow equals bad (Luoh and Tsaur 2011; Bae and Dae-Young 2014). the positive influence of age on taste perception, chef’s expertise. The premise of this research is that this notion may be incorrect. The study shows that the age effect is mitigated when people are Although older workers are indeed perceived to be slower, we argue informed that both chefs have high expertise. We demonstrate that that when it comes to food, slow is good. Similar to the time taken when the expertise on doing the food was framed as low the older by a craftsman to perfect a project, we argue that consumers have a chef lead to better taste perceptions and likelihood to visit the res- lay belief that “slow food=better food”. Grounded in the literature taurant, supporting our argument that consumers perceive the food about the Slow Food Movement, which advocates a more thoughtful prepared by an older chef as better. In contrast, when both chefs are consumption and preparation of food (Dunn et al. 2011), we argue framed as having high expertise the age no longer impacts consum- that in food context, slowness can be associated with effort and care. ers’ perceptions and attitudes. Therefore, because older people are slower this slow pace can lead to Through four studies—two involving actual taste experiences positive perceptions from consumers. and two employing scenarios —we show that employee age can pos- More formally, itively impact consumers’ taste perceptions and purchase attitudes. Consequently, this research makes a number of theoretical and mana- Hypothesis 1: When food is prepared by an older (younger) gerial contributions. From a theoretical perspective, the positive ef- person, consumers will evaluate it as more (less) fect of age on taste contributes to age stereotype literature and shows tasty. that there are contexts in which age is an advantage (Marcus et al. 2016). Our mediation of “slow food=better food” also contributes to Hypothesis 2: The effect of age on taste evaluation is mediated the literature on lay beliefs and food assessment (Hoegg and Alba by the perceived preparation time. 2007; Raghunathan, Naylor, and Hoyer 2006; Werle, Trendel, and Ardito 2013; Mai and Hoffman, 2015). Finally, this research also has With four studies we demonstrate that although consumers in- important managerial implications and shows that restaurant manag- deed think that the older cook takes longer to prepare food, they also ers should be more open to hiring older workers and should commu- believe the food prepared by her will be tastier than that prepared by nicate it to consumers. a younger person. Study 1 (N=648) was conducted online using MTurk. Each REFERENCES participant evaluated 2 products (from a total of 8 products investi- Bae, Gumkwang, and Dae-Young Kim (2014), “The effects of gated). For that participants saw a picture containing the photo of the offering menu information on perceived waiting time.” Journal product besides a photo of the producer (which was young or old), of Hospitality Marketing & Management, 23 (July) 746-767. and the information about the producer’s age. Congruently to the lit- DeArmond, Sarah, Mary Tye, Peter Y. Chen, Autumn Krauss, D. erature on age stereotype we show that people tend to judge non-food Apryl Rogers, and Emily Sintek. (2006), “Age and Gender related products (e.g. fashion accessories, electronics, decoration ser- Stereotypes: New Challenges in a Changing Workplace vices, nail polisher) more negatively when designed and developed and Workforce.” Journal of Applied Social Psychology, 36 by older people in comparison to their younger counterparts. How- (August), 2184-2214. ever, for food or beverage products (e.g. cookies, soup, cake, chicken Haws, K. L., Reczek, R. W., & Sample, K. L. (2016). Healthy and tea) we found an inverse and positive impact of older age. Diets Make Empty Wallets: The Healthy= Expensive In study 2, 109 undergrad students participated in a cupcake Intuition. Journal of Consumer Research, 43(6), 992-1007. tasting. The product was supposedly made by Tiane, the cook, who had either 25 or 55 years old. We manipulated age with two hired ac-

Hoegg, JoAndrea and Joseph W. Alba, (2007), “Taste perception: Ng, Thomas W. H. and Daniel C. Feldman (2012), “Evaluating six more than meets the tongue,” Journal of Consumer common stereotypes about older workers with meta‐analytical Research, 33 (March), 490-498. data,” Personnel Psychology, 65 (December), 821-858. Kaufmann, M Michèle C., Franciska Krings, and Sabine Sczesny Raghunathan, Rajagopal, Rebecca W. Naylor, and Wayne D. Hoyer (2015), “Looking Too Old? How an Older Age Appearance (2006), “The unhealthy= tasty intuition and its effects on taste Reduces Chances of Being Hired,” British Journal of inferences, enjoyment, and choice of food products,” Journal Management, 1-13. of Marketing, 70 (October), 170-184. Luoh, Hsiang-Fei and Sheng-Hshiung Tsaur (2011), “Customers’ Toossi, Mitra (2015), “Labor force projections to 2024: the labor Perceptions of Service Quality: Do Servers’ Age Stereotypes force is growing, but slowly,” Monthly Labor Review, 138 Matter?” International Journal of Hospitality Management, 30 (December). (June), 283-89. United Nations (2013), “World population ageing 2013”, available Mai, Robert and Steffan Hoffman (2015), “How to Combat the at: www.un.org/en/ development/desa/population/publications/ Unhealthy = Tasty Intuition: The Influencing Role of Health pdf/ageing/WorldPopulationAgeing2013.pdf (accessed June Consciousness,” Journal of Public Policy & Marketing, 34 (1), 05, 2016). 63-83. Werle, Carolina O., Olivier Trendel, and Gauthier Ardito (2013), Marcus, Justin, Barbara Ann Fritzsche, Huy Le, and Michael “Unhealthy food is not tastier for everybody: The “healthy= Dennis Reeves (2016), “Validation of the work-related tasty” French intuition,” Food Quality and Preference, 28 age-based stereotypes (WAS) scale,” Journal of Managerial (April), 116-121. Psychology, 31 (May), 989-1004. Back to the Golden Age: When Economic Insecurity Promotes Nostalgic Consumption Chun-Ming Yang, Ming Chuan University, Taiwan

EXTENDED ABSTRACT condition on the product choice index. Mean scores of the product Although recent market studies suggest that consumers may choice index were higher for the high-PEI condition compared with prefer nostalgic products and services when facing economic uncer- the low-PEI condition. The results were reversed for meaning in life. tainty, no empirical effort has been made to verify the influence of A bootstrap test again confirmed the indirect path from the PEI con- perceived economic insecurity (PEI) on consumers’ nostalgic condition through meaning in life to the product choice index. sumption and to identify which psychological mechanism is activat- Study 3 was conducted to conceptually replicate and extend the ed in response to PEI. Therefore, we attempted to fill this gap in this findings from study 2. First, we ascertained whether our experimen- study. Building on previous research on nostalgia and meaning in life, tal findings can be replicated in populations facing various levels of we suggest that PEI, both chronic and temporary, lowers consumers’ chronic PEI. Specifically, rather than manipulating the momentary meaning in life, which in turn makes consumers feel nostalgic, prefer sense of economic insecurity, we treated PEI as an individual differ- nostalgic products, and respond favorably to nostalgic ads. ence factor and tested how this variable influences consumer prefer- Why does PEI affect nostalgic consumption? Prior studies re- ence toward nostalgic products. Second, we used popular music as port that economic uncertainty affects consumers’ various responses the product context to increase the generalizability of our findings. such as prosocial behaviors, preference polarization, investment In total, 149 adults participated in this study. As predicted, PEI had in offspring, saving behavior, health care decisions, creativity, and a positive effect on the number of nostalgic songs chosen by par- preferences toward scarce goods. Recently, evidence indicates that ticipants but had a negative effect on meaning in life. Bootstrapping PEI also lowers consumers’ meaning in life (Abeyta et al. 2016), analysis revealed that meaning in life mediated the relationship be- because PEI threatens the continued satisfaction of basic psycho- tween PEI and the number of nostalgic songs selected. logical needs and therefore hinders consumers’ abilities to pursue In study 4, we tested the meaning-in-life account by examining personal meaningful goals. A restorative perspective highlights that a theoretically relevant moderator: social connectedness (SC). Cur- nostalgia can counteract insecurities and threats to meaning in life by rent research suggests that enhanced SC may alleviate the effects of providing a reservoir of meaningful personal experiences that help PEI on meaning in life. In total, 184 adults participated in this study. ease the discomfort (Routledge et al. 2011; Zhou et al. 2008). For Upon arrival, the participants were informed that the study involved example, Loveland et al. (2010) found that social exclusion increases three unrelated parts. The first part was a general “Lifestyle Survey” consumers’ preference toward nostalgic products, and consumption that measured participants’ childhood socioeconomic status and SC. of nostalgic products reduces the social discomfort. By manipulating The second part included the same PEI manipulation used in study death awareness, Routledge et al. (2008) revealed that nostalgia can 2. In the final part, participants received a full-color nostalgic ad buffer the effects of existential threat on death-thought accessibility. promoting consumers to recall “the good times” of their past. After To summarize, we hypothesize that PEI, both temporary and chronic, exposure to the ad, participants were asked to report their purchase increases consumers’ preference toward nostalgic consumption. This intention. We found significant interactions between PEI and SC on is because PEI threatens consumers’ meaning in life. To counteract both meaning in life and purchase intention. Subsequent spotlight the threat, nostalgic consumption can act as an “inner refuge” and analysis revealed that low-SC respondents were affected by the PEI provide people with psychological shield and warmth. Four studies condition. By contrast, high-SC participants were less affected by tested this hypothesis. the PEI condition. We also uncovered a conditional indirect effect of In study 1, 174 adults were recruited to complete an online the interaction between PEI and SC on purchase intention through questionnaire. We measured participant’s PEI, nostalgic index (Hol- meaning in life. The index of moderated mediation for the indirect brook, 1994), meaning in life, childhood socioeconomic status, age, effect was significant. and current household income. The results indicated that higher PEI From a theoretical perspective, this research is the first to show was associated with lower meaning in life and a higher nostalgia in- that PEI, a threat to meaning in life, can elicit favorable consumer dex. A 5000 resample bootstrap provided evidence supporting this responses toward nostalgic products and ads. This article also con- indirect effect. In all four studies, childhood socioeconomic status tributes to the broader nostalgia field. The literature on nostalgia does had a negative impact on consumer’s nostalgic responses, but had no not consider the influence of economic factors. The results of our significant interaction effect with PEI. research shed new light on the effects of insecurity on meaning in life Study 2 examined whether PEI manipulation increases interest and nostalgia. Our findings also have useful implications for market- in consuming nostalgic products. Additionally, study 2 tested the me- ing practitioners; that is, adding nostalgic elements into marketing diating role of meaning in life in driving the preference for nostalgic activities may substantially improve consumers’ reactions. products after PEI manipulation. In total, 119 adults were randomly assigned to either the high- or low-PEI condition. In the first task, we REFERENCES used a news article to manipulate participants’ PEI. In the high-PEI Abeyta, Andrew A., Clay Routledge, Michael Kersten, and Cathy condition, the article described the recent economic recession and R. Cox (2016), “The Existential Cost of Economic Insecurity: how the economic problems are likely to persist into the subsequent Threatened Financial Security Undercuts Meaning,” Journal of year. In the low-PEI condition, the participants read an article similar Social Psychology, DOI: 10.1080/00224545.2016.1270892 in length to the high-PEI condition, which described a recent air pol- Loveland, Katherine E., Dirk Smeesters, and Naomi Mandel lution problem caused by a major manufacturer. After reading the (2010), “Still Preoccupied with 1995: The Need to Belong article, participants finished relevant measurement. Subsequently, and Preference for Nostalgic Products,” Journal of Consumer participants completed five choices between nostalgic and nonnos- Research, 37 (October), 393–408. talgic brands in various product categories to form a nostalgic product choice index. Results showed a significant main effect of the PEI

Routledge, Clay, Jamie Arndt, Constantine Sedikides, and Tim Zhou, Xinyue, Constantine Sedikides, Tim Wildschut, and Wildschut (2008), “A Blast from the Past: The Terror Ding-Guo Gao (2008), “Counteracting Loneliness: On the Management Function of Nostalgia,” Journal of Experimental Restorative Function of Nostalgia,” Psychological Science, 19 Social Psychology, 44 (January), 132–140. (10), 1023–1029. Routledge, Clay, Jamie Arndt, Tim Wildschut, Constantine Sedikides, Claire M. Hart, Jacob Juhl, Ad J. J. M. Vingerhoets, and Wolff Schlotz (2011), “The Past Makes the Present Meaningful: Nostalgia as an Existential Resource,” Journal of Personality and Social Psychology, 101(3), 638-652. How Group Identification and Stereotype Content Determine the Effectiveness of Ads Portraying Positive Stereotypes Linyun W. Yang, University of North Carolina at Charlotte, USA Tanya L. Chartrand, Duke University, USA Gavan J. Fitzsimons, Duke University, USA

EXTENDED ABSTRACT respond more favorably to ads portraying competence-related ste- While prior consumer behavior research on stereotypes has large- reotypes. ly focused on negative stereotypes (e.g., Lee, Kim and Vohs 2011), this An online panel of Asian Americans completed two ostensibly research examines how positive stereotypes portrayed in advertising unrelated questionnaires. In the first questionnaire, we manipulated influence the effectiveness of these ads for consumers belonging to the group identification by having participants respond to three open- stereotyped group. We propose that how consumers respond depends ended questions regarding their feelings about being Asian adapted on their identification with the stereotyped group and the content of from Leach et al. (2008). Then participants rated an ad for “Sher- the positive stereotype (warmth vs. competence). We suggest that for wood Test Prep,” a company that specialized in helping students im- high identifiers, positive stereotypes in advertising elicit a communal prove their quantitative skills. While the stereotypical ad featured an orientation and an increased valuation of warmth information, whereas Asian American spokesperson, the control ad featured a non-Asian for low identifiers, these stereotypes cue an agentic orientation and an American spokesperson. In addition to the same evaluation items increased valuation of competence information. used in the Study 1, participants provided product ratings (useful- We propose that the effectiveness and persuasiveness of an ad ness, value, effectiveness, purchase interest; α=.96). portraying positive stereotypes lies in how much consumers value An ANOVA on product evaluations revealed a two-way ad type the ad’s arguments, which are often closely tied to the portrayed ste- by group identification prime interaction (F(3, 68)=3.72, p=.05), reotype (i.e., warmth vs. competence). When a positive stereotype is showing that those primed with high relative to low identification portrayed in advertising, the brand and promoted product becomes provided less favorable product evaluations, but only after viewing linked to characteristics tied to the stereotype. Consequently, we the Asian stereotype and not the control ad. predict that high identifiers will respond more favorably than low In Study 3, we examine the responses of Asian American wom- identifiers to ads that stereotype their group as warm because high en who are unique in that they can be positively stereotyped as wom- identifiers are more likely to have a communal orientation and there- en along the warmth dimension or as Asian along the competence fore be more persuaded by warmth-related arguments. In contrast, dimension (Shih, et al., 1999). We test the prediction that high (vs. low identifiers will respond more favorably than high identifiers to low) identifying Asian American women will respond more favor- ads stereotyping their group as competent because low identifiers ably to an ad that positively stereotypes women as empathetic and are more likely to adopt an agentic orientation and find competence- cooperative but less favorably to an ad that positively stereotypes related arguments more useful and informative. Asians as intelligent and quantitatively skilled. In Study 1, we first focus on American Southerners to test the An online panel of Asian American women completed both hypothesis that ads portraying positive stereotypes related to warmth gender and ethnicity versions of Leach et al.’s (2008) scale in ran- will be more effective for those who identify highly with the stereo- dom order one week prior. For the main study, participants viewed typed group. We also tested whether an increased communal orienta- an ad featuring an Asian American woman promoting a company tion for high relative to low identifiers drives more favorable evalua- called “The Negotiation Institute” that positively stereotyped either tions of the promoted restaurant. women or Asian Americans and rated the company (good, favorable; Undergraduate students self-identified as Southerners complet- r=.65, p<.0001). ed Leach et al.’s (2008) measure of group identification one week in To test our predictions, we used two regression models to exam- advance. In the main study, participants viewed either a stereotypical ine product evaluations. For the ad type x female group identification ad featuring Southerners promoting a restaurant that provided warm model, analysis revealed a significant two-way interaction (t(71)=- and welcoming service or a control ad which was identical except 2.20, p<.05). Those high relative to low in female group identifica- with no information regarding the spokespeople’s origins. Partici- tion indicated more favorable evaluations after viewing the female pants then rated the restaurant (good, favorable; r=.72, p<.0001) and stereotype ad but not after viewing the Asian stereotype ad. For the the importance of communal values (e.g., altruism, harmony, Trap- ad type x Asian American identification model, we also found a sig- nell and Paulhus 2012). nificant two-way interaction (t(71)=-3.14, p<.01). For the Asian ste- Regression analysis revealed a two-way interaction between ad reotype ad, those low relative to high in Asian American identifica- type and group identification (t(99)=-2.42, p<.05). For the South- tion provided more favorable evaluations. For the female stereotype ern stereotype ad, high relative to low identifiers indicated more ad, those low versus high in Asian American identification actually favorable evaluations, but this was not the case for the control ad. evaluated the company less favorably. The same analysis on communal values endorsement also yielded In conclusion, as the marketplace becomes more diverse, it is a two-way interaction (t(99)=-2.04, p<.05) and same pattern of ef- increasingly important to understand how consumers respond to fects. Moderated mediation analysis confirmed that communal val- portrayals of their social group. Given their complimentary nature, ues endorsement mediated the positive effect of group identification positive stereotypes are frequently and oftentimes inadvertently por- on evaluations when participants viewed the Southern stereotype ad trayed in the media, highlighting the need for more research on topics (B=.11, 95% CI: .01, .32), but not when participants viewed the con- in this area. We hope the current research is useful for several dif- trol ad (B=-.01, 95% CI: -.11, .04). ferent audiences, including marketers appealing to diverse markets, In Study 2, we examine ads that positively stereotype Asian stereotyped consumers sorting through their own responses, and re- Americans as being intelligent and quantitatively skilled (Lin et al., searchers exploring the psychological processes elicited by positive 2005) to test the prediction that low relative to high identifiers will stereotypes in the marketplace.

Advances in Consumer Research 974 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 975 REFERENCES Lin, Monica H., Virgina S. Y. Kwan, Anna Cheung, and Susan T. Leach, Colin W., Martijn van Zomeren, Sfven Zebel, Michael Fiske (2005), “Stereotype Content Model Explains Prejudice L. W. Vliek, Sjoerd F. Pennekamp, Bertjan Doosje, Jaap for an Envied Outgroup: Scale of Anti-Asian American W. Ouwerkerk, and Russell Spears (2008), “Group-Level Stereotypes,” Personality and Social Psychology Bulletin, 31 Self-Definition and Self-Investment: A Hierarchical (1), 34 – 47. (Multicomponent) Model of in-Group Identification,”Journal Shih, Margaret, Todd L. Pittinsky, and Nalini Ambady (1999), of Personality and Social Psychology, 95 (1), 144-65. “Stereotype Susceptibility: Identity Salience and Shifts in Lee, Kyoungmi, Hakkyun Kim, and Kathleen D. Vohs (2011), Quantitative Performance,” Psychological Science, 10 (1), “Stereotype Threat in the Marketplace: Consumer Anxiety and 80-3. Purchase Intentions,” Journal of Consumer Research, 38 (2), Trapnell, Paul D. and Delroy L. Paulhus (2012), “Agentic and 343-57. Communal Values: Their Scope and Measurement,” Journal of Personality Assessment, 94(1), 39-52. Effects of Matched and Mismatched Comparative Advertising Messages: The Moderating Role of Consumers’ Processing Focus Xiaojing Yang, University of Wisconsin - Milwaukee, USA Shailendra Jain, University of Washington, USA Charles Lindsey, State University of New York, Buffalo, USA Frank Kardes, University of Cincinnati, USA

EXTENDED ABSTRACT whereas when it comprises a relatively large set of brands, consum- A frequently employed framework to account for findings in ers may be more likely to hold prototypical representation of the cat- persuasion research focuses on whether persuasive messages match egory (e.g., Yeung and Soman 2007). Given that both comparative or mismatch consumers’ self-schemata (e.g., extroversion; Wheeler, messages of different types (“leading brand” and “other brands”) and Petty, and Bizer 2005), goals (e.g., regulatory focus; Lee, Keller, and categories of differing sizes may influence the reliance on prototype- Sternthal 2010), emotions (Agrawal and Duhachek 2010), or contex- vs. exemplar-based processing, conceptually, and from a research tual factors during message processing (e.g., temporal distance; Zhao design viewpoint, a match exists when the perceived category size is and Xie 2011). Most of this research has documented a “matching” large (vs. small) and consumers are exposed to an “other brands” (vs. effect of advertising messages, suggesting that matched messages are a “leading brand”) message. This is so because both such conditions more persuasive than mismatched messages, though on some (but promote a prototype-(vs. exemplar-) based representation. rare) occasions, mismatched messages have been found to be more On the basis of metacognition research (Alter et al. 2007) and effective (Millar and Millar 1990; Aaker and Williams 1998). the finding that matched messages are easier to process than mis- Researchers have recently set out to account for these seem- matched messages, we contend that when processing matched ingly opposing findings of matched and mismatched messages. For comparative advertising messages (“leading brand” claims in a example, Malaviya and Brendl (2014) show that hedonic motives small category or “other brands” claims in a large category), con- moderate the effects of matched vs. mismatched messages. They find sumers should experience less processing difficulty, which in turn that favorable persuasion outcomes from a match (vs. mismatch) be- will engender a feeling-based processing strategy that is affective, tween regulatory focus and message framing occur only when the intuitive, and holistic; however, exposure to mismatched compara- hedonic motive of the perceiver is consistent with (vs. opposes) that tive messages (“leading brand” claims in a large category or “other of the message. Extending this stream of research, we examine the brands” claims in a small category) will lead consumers to resort to moderating role of consumers’ processing focus (cognitive vs. af- a more analytical processing strategy that is cognitive, rational, and fective) and posit that consumers with a cognitive (vs. an affective) deliberative. We further propose that the extent to which matched or focus find mismatched (vs. matched) messages more persuasive. mismatched comparative advertising messages elicit more favorable We conduct our inquiry by focusing on two types of indirect consumer responses hinges on consumers’ processing focus, whether comparative messages: “leading brand” comparisons (“brand X an- activated prior to ad exposure or chronically preferred by consumers. algesic relieves pain better than the leading brand”) and compari- That is, consumers should develop more favorable reactions toward sons featuring “other brands” (“brand X trucks are better than other mismatched comparative advertising messages if their processing fo- brands in their class”). We choose comparative advertising and, more cus is also more cognitive. If their processing focus is affective, they specifically, indirect comparative messages, as our context because: should develop more positive responses toward matched compara- (1) more than 70% of all advertising in the US may feature some tive advertising messages. type of comparison (Pechmann and Stewart 1990), and (2) though We report two experiments which provide support to our indirect comparative advertising is significantly more prevalent in theorizing. In the first experiment, respondents were first exposed practice than direct comparative advertising (Pechmann and Stewart to a cognitive (or an affective) processing focus manipulation task 1990), most comparative investigations have focused on direct com- where they were asked to search words related to cognition or affect. parisons (e.g., Jain, Agrawal, and Maheswaran 2006; Zhang, Kardes, Next, they received information about the car market where category and Cronley 2000). size was manipulated (emphasizing there are many or few choices in Drawing on categorization research, we posit that different the car market). Then they viewed a car ad which featured the “lead- comparative claims and ad contexts may induce alternative catego- ing brand” claim or the “other brands” claim. Finally, they reported rization approaches and distinct types of elaborations. First, consis- their brand evaluations. Our results find that when a cognitive (vs. an tent with Snyder (1992), we posit that “leading brand” claims invite affective) focus was induced, respondents reported more favorable consumers to compare the advertised brand to a specific (albeit un- brand evaluations where there was mismatch (vs. match) between named) brand, leading to exemplar-based processing. Further, “other category size and type of claim. In the second experiment, the cog- brands” claims invite consumers to compare the advertised brand to nitive/affective focus was operationalized using chronic construal- a group of brands, eliciting prototype-based processing. In addition, level, which was measured using the behavioral identification form. we propose that when a category consists of relatively few brands, Findings from experiment 1 were replicated. consumers may be more predisposed to exemplar-based processing

Advances in Consumer Research 976 Volume 45, ©2017 Regulatory Focus and Donation to Proactive and Reactive Causes Zhiyong Yang, University of Texas at Arlington, USA Sijie Sun, University of Texas at Arlington, USA Narayan Janakiraman, University of Texas at Arlington, USA Ritesh Saini, University of Texas at Arlington, USA

EXTENDED ABSTRACT to reactive ones, because of a latent expectation that they or their off- Charitable organizations often adopt divergent approaches to spring may potentially benefit from the former, but not so much from tackle the same problem. The American Heart Association (AHA), the latter. However, individuals with a prevention focus are equally for example, has two distinct types of programs to deal with diabe- likely to donate to both, as their tendency to donate is mainly driven tes: (1) one aiming at fun research to develop new treatments and by obligations and responsibilities, which makes cause type less rel- prevention strategies, and (2) the other focusing on assisting current evant. These propositions find resonance in recent research exam- diabetes patients pay for their treatment. How do potential donors ining the impact of regulatory focus on green consumption (Ku et react to these two types of causes? Why? This is the question that this al. 2012). Although not in the donation context, these findings show research intends to address. that prevention-focused consumers are more strongly persuaded by In this paper we propose a duality in altruistic motivations. environment-benefit appeals, whereas promotion-focused consum- While one pro-social motivation drives support for proactive causes ers are more responsive to self-benefit appeals. This suggests that where the focus is on preventing negative future outcomes, the other individuals with a promotion focus are more likely to focus on ways motivation impels support for reactive actions where the focus is on and means to self-benefit and be self-sufficient, while individuals mitigating existing negative situations. While AHA’s initiatives for with a prevention focus should be other-focused given their drive to research and prevention are proactive, its help to existing patients fulfill their obligations and responsibilities to others. The forgoing is reactive. A proactive cause refers to the fund-raising programs discussion suggests that perceived self-benefit is the key mechanism that aim to prevent something bad from happening before it occurs, through which regulatory focus affects individuals’ preference to whereas reactive cause is characterized by the remedies that aim to proactive versus reactive causes. resolve the problem after it arises. Notably, this classification is in If our theorization is correct, when other-benefit of the cause is line with the service marketing literature, which shows that there are made salient, social desirability of helping gets activated and self- two possible strategies to deal with service failure, proactive and re- serving motivations are subdued (Fisher et al. 2008). This will lead active. Proactive strategies are implemented before service failure, promotion -focused individuals to behave similarly to prevention- whilst reactive strategies operate after a service problem appears focused individuals. (Worsfold et al. 2007). From a conceptual perspective, the two types The proposed effects are tested in three studies. Study 1 uses of donation causes differ in who the beneficiaries of the donation are chronic measures to test the core thesis of this research, i.e., promo- likely to be. While a proactive cause may potentially benefit them- tion focus leads to more preference for proactive (vs. reactive) causes selves (including one’s offspring, and future generations), a reactive but prevention focus does not differentiate between these two types cause narrowly focuses on the victims who are currently suffering of causes. Study 2 replicates study 1’s findings using situationally from the problem. activated regulatory focus, and provides direct evidence on the per- In this research, we predict that individuals’ regulatory focus has ceived self-benefit of the cause as the key mechanism. Study 3 ma- differential effects on their preference for these two types of causes. nipulates the salience of self-benefit and other-benefit for the same According to Higgins (1998), individuals possess two types of regu- cause, and shows that when self-benefit of the cause is made salient, latory goals, namely promotion-focused and prevention-focused. the findings of studies 1 and 2 are replicated; however, when other- The goals of promotion-focused individuals are characterized by the benefit of the cause is made salient, individuals with a promotion- orientation toward seeking benefits, gains, and self-enhancements, focus do not differentiates the causes. whereas prevention-focused people are oriented toward avoiding Study 1 (N=304) was conducted online using MTurk. Partici- uncertainty, mistakes, and losses (Aaker and Lee 2001). Because of pants first read the description of AHA then were asked to indicate this, individuals with a promotion focus tend to eagerly pursue their their preference for these two types of causes (1 = definitely prefer own interest and desires without much consideration for possible the causes that help the patients deal with their current financial social consequences (Brebels et al. 2008). In contrast, those with a needs; 7 = definitely prefer the causes that aim at preventing diseas- prevention focus are directed toward goals like duties, responsibili- es from happening in the future). Afterwards, participants provided ties, and obligations, and are more considerate to possible social con- ratings on their level of regulator foci, using Haws et al.’s (2010) sequences that represent a threat to things one ought to do (Brebels scale. Participants’ age, gender, and family history of heart disease et al. 2008; Oyserman et al. 2007). Consistent with this qualitative was measured as covariates. A regression analysis with preference of difference across promotion- and prevention-focus, cross-cultural lit- donation cause as the criterion and prevention focus and promotion erature shows that the goal of individualistic countries (e.g., the U.S. focus as independent variables revealed a significant positive effect and Canada, where people are usually promotion-focused) is to de- of promotion focus on preference (β = .14, t(301) = 2.36, p < .05). velop an individual sense of identity and self-sufficiency away from The effect of prevention focus on preference was not significant (β = others (Triandis 1995). In contrast, from an early age, individuals in -.08, t(301) = -1.45, p = .15). None of the control variables reached collectivistic cultures (e.g., China and India, where people are usu- statistical significance (all p’s > .50). These results support our hy- ally prevention-focused) are trained to become an integral part of the pothesis that individuals with a promotion focus prefer the proactive larger group and make contributions to the achievement and welfare cause over reactive cause, whereas individuals with a prevention fo- of the collective (Triandis 1995). cus have no difference toward the two causes. Based on such differences, we expect that individuals with a The purpose of study 2 (N=172) is threefold: (1) to replicate promotion focus are more likely to donate to proactive causes than study 1’s finding through manipulating regulatory foci, (2) - tofor

This research contributes to the literature in three important mally test perceived self-benefit of the cause as the key mediator ways. First, most studies in this domain focus on donation in gen- underlying our results, and (3) to rule out long-term orientation as eral, without differentiating the cause type. Recently, non-profit -or an alternative explanation. It featured a 2 (regulatory foci: preven- ganizations have started to offer more than one project for the same tion vs. promotion) × 2 (cause type: proactive vs. reactive) between- issue. In this research, we provide theoretical explanation about the subjects design. Regulatory foci were manipulated the same way as difference between two types of causes, namely proactive and reac- other researchers (Mourali et al. 2007). After the regulatory foci ma- tive. We suggest that these two causes may induce different levels nipulation, participants were provided with an appeal from a charita- of perceived self-interest and therefore affect individuals’ donation ble organization, framed as either proactive or reactive. Participants behavior. Such conceptual distinctions among donation causes open were asked how much they would like to donate to this project. The a new avenue for future research. Second, our research also extends mediatior, perceived self-benefit associated with the cause, was mea- the literature on regulatory focus. Previous studies on regulatory foci sured by two scales from White and Peloza (2009). A 2 (regulatory have mainly focus on the distinctions across risk perception. Extend- foci) × 2 (cause type) ANOVA on the donation amount found only ing this stream of research, the present study shows that regulatory a significant regulatory focus × cause type 2-way interaction (F(1, foci also drive individuals to be sensitive to self-benefit embedded 167) = 5.80, p < .05). Neither of the main effects was significant into the decision object. Specifically, individuals with a promotion (p’s > .50). When primed with promotion focus, individuals donated focus are more responsive to self-benefit compared to individuals more (M = 12.44) to the proactive cause than to the reactive cause with a prevention focus. These findings also add to our understand- (M = 4.82, t(83) = -2.00, p < .05). In contrast, when primed with ing of self-benefit/other benefit appeals. For example, prior research prevention focus, the donation amount did not vary across the two indicates that other-benefit fund-raising appeals are more effective causes (M = 6.82, M = 11.70; t(82)= 1.09, p = .28). These proactive reactive compared to self-benefit appeals (Fisher et al. 2008). Our results results replicated study 1’s findings. Mediation was assessed with qualify these findings by demonstrating that this is not true for pro- the bootstrapping method, using Hayes’s (2012) PROCESS macro. motion-focused individuals. When other-benefit is salient, promo- Consistent with our predictions, the model indicates that the indirect tion-focused individuals perceive the two donation causes equally. effect of moderated mediation is significant (95% CI = .15 to 7.66). When self-benefit is salient, promotion-focused individuals will do- A follow-up study (N=86) was conducted using a 2 (regulatory nate more to proactive causes. foci: prevention vs. promotion) × 2 (donation cause: proactive vs. reactive) mixed design, with regulatory foci being a between-subjects REFERENCE factor and donation cause as a within-subjects factor. Regulatory fo- Aaker, Jennifer L. and Angela Y. Lee (2001), “‘I’ seek pleasures cus manipulation and the donation causes were the same as in study and ‘we’ avoid pains: The role of self-regulatory goals in 2, except that participants viewed two causes and choose only one information processing and persuasion,” Journal of Consumer of them to donate. A binary logistic regression on the choice of pro- Research, 28(1), 33-49. active (vs. reactive) cause showed a significant effect of regulatory Bearden, William O., R. Bruce Money, and Jennifer L. Nevins foci on decision outcome (β = .89, p < .05). For individuals primed (2006), “A measure of long-term orientation: Development with promotion focus, a greater portion selected the proactive cause and validation,” Journal of the Academy of Marketing Science, (63.9%) compared to the reactive cause (42.0%, z = 2.00, p < .05). 34(3), 456-467. In contrast, for individuals primed with prevention focus, a smaller Brebels, L., De Cremer, D., and Sedikides, C (2008), “Retaliation portion selected the proactive cause (36.1%) compared to the reac- as a response to procedural unfairness: A self-regulatory tive cause (58.0%, z = -2.00, p < .05). approach,” Journal of Personality and Social Psychology, The purpose of study 3 (N=158) is to test the moderation effect 95(6), 1511–1525. of the salience of benefit associated with the cause (self vs. other). Fisher, Robert J., Mark Vandenbosch, and Kersi D. Antia (2008), Since prevention focus leads to non-differentiation between the two “An empathy-helping perspective on consumers’ responses to types of donation causes, we only focus on promotion focus in this fund-raising appeals,” Journal of Consumer Research, 35(3), study. We expect that in self-benefit conditions, individuals with a 519-531. promotion focus intend to donate more to proactive causes than to Hamstra, Melvyn RW, Nico W. Van Yperen, Barbara Wisse, and Kai reactive causes; however, in other-benefit conditions, they do not Sassenberg (2011), “Transformational-transactional leadership differentiate between these two causes. All participants were first styles and followers’ regulatory focus,” Journal of Personnel manipulated by promotion focus, using the same stimuli as used in Psychology, 10(4), 182-186. study 2. They were then randomly assigned to the conditions in a Haws, Kelly L., Utpal M. Dholakia, and William O. Bearden 2 (cause type: proactive vs. reactive) × 2 (salience of benefit: self- (2010), “An assessment of chronic regulatory focus measures,” benefit vs. other-benefit) between-subjects design. Donation cause Journal of Marketing Research, 47(5), 967-982. was manipulated in the same way as in study 2, except that an ad- Hayes, Andrew F (2012), “PROCESS: A Versatile Computational ditional sentence was added before the donation intention measure Tool for Observed Variable Mediation, Moderation, and to manipulate self-benefit [other-benefit] salience. We performed a Conditional Process Modeling,” Columbus: Ohio State 2 (cause type) × 2 (salience of benefit) ANOVA with the donation University. http://www.afhayes.com/public/process2012.pdf. amount as the dependent variable. Neither main effects was signifi- Higgins, E. Tory (1998), “Promotion and prevention: Regulatory cant (p’s > .15). More important and pertinent to our hypothesis, focus as a motivational principle,” Advances in Experimental there was a significant perceived benefit × cause type 2-way interac- Social Psychology, 30, 1-46. tion (F(1, 152) = 6.85, p = .01). When self-benefit was made salient, Ku, Hsuan-Hsuan, Chien-Chih Kuo, Ching-Luen Wu, and Chih- individuals with a promotion focus donated more (M = 17.95) to the Ying Wu (2012), “Communicating green marketing appeals proactive cause than to the reactive cause (M = 11.44, t(82) = -2.07, effectively: The role of consumers’ motivational orientation to p < .05). In contrast, when other-benefit was made salient, the dona- promotion versus prevention,” Journal of Advertising, 41(4), tion amount did not vary across two causes (M = 9.74, M = proactive reactive 41-50. 13.53; t(72) = 1.46, p = .15). Advances in Consumer Research (Volume 45) / 979

Mourali, Mehdi, Ulf Böckenholt, and Michel Laroche (2007), White, Katherine, and John Peloza (2009), “Self-benefit versus “Compromise and attraction effects under prevention and other-benefit marketing appeals: Their effectiveness in promotion motivations,” Journal of Consumer Research, generating charitable support,” Journal of Marketing, 73(4), 34(2), 234-247. 109-124. Oyserman, Daphna, Ayse K. Uskul, Nicholas Yoder, Randy M. Worsfold, Kate, Jennifer Worsfold, and Graham Bradley (2007), Nesse, and David R. Williams (2007). “Unfair treatment “Interactive effects of proactive and reactive service recovery and self-regulatory focus,” Journal of Experimental Social strategies: The case of rapport and compensation,” Journal of Psychology, 43(3), 505–512. Applied Social Psychology, 37(11), 2496-2517. Triandis, Harry Charalambos (1995), Individualism & Collectivism, Westview press. The Paradoxical Impact of Positive Mood on Goal Favorability Haiyang Yang, Johns Hopkins University, USA Antonios Stamatogiannakis, IE University, Spain

EXTENDED ABSTRACT website for one year. Random-effect logistic regression models This research examines the interplay of consumers’ positive showed that, consistent with our proposition, consumers in a mood and the type of goals they pursue. We propose and demonstrate chronically more (vs. less) positive affective state were more likely that although consumers believe that attainment goals are better for to set a weight-maintenance than weight-reduction goal. fostering a positive mood, they actually favor maintenance goals Study 2 had a 2 (positive vs. negative mood prime) × 2 more than attainment goals when they are in a positive mood. (maintenance vs. attainment goals) between-participant design, and Attainment and maintenance goals are two common types of utilized goal domains that participants self-identified as important goals (Yang, Stamatogiannakis, and Chattopadhyay 2015). For (e.g., work, health). A significant interaction effect revealed that, attainment goals, the current state of consumers is different from under positive mood, participants favored maintenance goals more their desired state (e.g., save, at least, $10 more this month). For than attainment goals. However, under negative mood, no difference maintenance goals, the current state already matches the desired was found. state; consumers pursuing this type of goal seek to maintain a current Study 3 had a 3 (neutral mood vs. positive mood vs. positive mood state that is the same as or exceeds the desired state (e.g., keep saving, and cognitive load) × 2 (health goal: attainment vs. maintenance) at least, the same amount this month). between-participant design. A significant interaction effect revealed Because attainment goals have better outcomes than the that maintenance goals were favored more than attainment goals corresponding maintenance goals (e.g., save $10 more vs. the same), in both positive mood conditions (showing qualitatively identical the outcomes of the former would be considered more affectively results). However, in the neutral mood condition, no difference was positive (cf. Plemmons and Weiss 2013). As such, consumers who found. Thus, consistent with our proposed nonconscious process, the seek to foster and elevate their positive mood may favor attainment effect persisted when conscious cognitive resources were constrained. goals over maintenance ones (Wegener et al. 1995). This reasoning is Study 4 had a 2 (positive vs. neutral mood prime) × 2 (health consistent with consumers’ lay beliefs shown in a pilot study: When goal: maintenance vs. attainment) between-participant design and given a choice between the two goal types, participants strongly measured participants’ chronic awareness of factors influencing their preferred attainment over maintenance goals in order to maintain mood. The mood × goal type interaction effect was replicated. More a positive mood. That is, according to extant mood maintenance importantly, the results showed a significant 3-way interaction: The theories as well as consumers’ lay beliefs shown in the pilot less participants were aware of factors influencing their mood, the study, consumers may be more favorable towards attainment (vs. more they favored maintenance (vs. attainment) goals under positive maintenance) goals when they are in a positive mood. mood, suggesting that the effect is nonconscious. However, in this research, we argue and show that the opposite Overall, our research contributes to understanding the interplay pattern can occur. We build on the prior research finding that between mood and goal pursuit. Although extant mood maintenance positive affect can be generated by pursuing other goals involving theories and consumers’ lay beliefs predict that attainment goals activities that are congruent with a target goal (cf. Kruglanski would be better for fostering a positive mood, we propose and 2006). For example, when team members’ private goals involve show in the field and lab that consumers in a positive (as opposed to activities that are perceived as congruent with the target goal they negative or neutral) mood tend to favor maintenance (vs. attainment) pursue for their team, they tend to experience more positive affect goals more. Our findings thus add to the current understanding of (Kristof-Brown and Stevens 2001). In our research context, positive the mood-goal relationship, complementing the existing theories mood maintenance can be considered a target goal consumers hold on mood maintenance (Handley et al. 2004; Wegener et al. 1995). (Wegener et al. 1995); attainment vs. maintenance goals (e.g., save Our results also complement the prior research finding that goal $10 more vs. the same) are the other goals that consumers pursue achievement generates positive affect (Plemmons and Weiss 2013) while they are after positive mood maintenance. Because consumers by showing that different affective states can have unique effects on tend to construe maintenance (but not attainment) goals as involving perceived favorability of different types of goals. Further, we also activities related to fostering and keeping one’s current state (Yang add to the literature by showing that this effect, resulting from the et al. 2015), maintenance (vs. attainment) goals are more congruent congruence between maintenance goals and the goal of positive with the target goal of maintaining positive mood. Thus, to the extent mood maintenance, is largely nonconscious. The effect persists that this congruence can generate positive affect (Kruglanski 2006), when consumers’ conscious cognitive resources are constrained, consumers in a positive (but not negative or neutral) mood would and is attenuated among consumers who are more aware of factors favor maintenance goals more than attainment goals. influencing their mood. Further, because both congruence effects (Kruglanski 2006) and mood maintenance (Handley et al., 2004) tend to operate REFERENCES nonconsciously, our propose interaction effect is also likely Handley, Ian M., G. Daniel Lassiter, Elizabeth F. Nickell, and nonconscious in nature. Thus, the effect would persist when Lisa M. Herchenroeder (2004), “Affect and Automatic Mood conscious cognitive resources are constrained, but be attenuated Maintenance,” Journal of Experimental Social Psychology, 40 among individuals more aware of factors influencing their mood. (1), 106-12. Four field and lab studies involving different goal pursuit domains Kristof-Brown, Amy L., and Cynthia Kay Stevens (2001), “Goal and different operationalizations of positive mood, support these Congruence in Project Teams: Does the Fit between Members’ propositions. Personal Mastery and Performance Goals Matter?,” Journal of Study 1 examined real-life, bodyweight-management goal- Applied Psychology, 86(6), 1083-95. setting behavior of over 1,300 consumers on a goal-management

Kruglanski, Arie W. (2006), “The Nature of Fit and the Origins of Wegener, Duane T., Richard E. Petty, and Stephen M. Smith (1996), “Feeling Right”: A Goal-Systemic Perspective,” Journal of “Positive Mood can Increase or Decrease Message Scrutiny: Marketing Research, 43(1), 11-14. The Hedonic Contingency View of Mood and Message Plemmons, Stefanie A., and Howard M. Weiss (2013), “Goals Processing,” Journal of Personality and Social Psychology, 69 and Affect” in A. A. Locke and G. P. Latham (Eds.),New (1), 5-15. Developments in Goal Setting and Task Performance, New Yang, Haiyang, Antonios Stamatogiannakis, and Amitava York: Routledge. Chattopadhyay (2015), “Pursuing Attainment versus Maintenance Goals: The Interplay of Self-Construal and Goal Type on Consumer Motivation,” Journal of Consumer Research, 42, 93-108. The “Smile-Seeking” Giver: How Immediate Affective Reactions Motivate and Reward Gift-giving Adelle Yang, National University of Singapore, Singapore Oleg Urminsky, University of Chicago, USA

EXTENDED ABSTRACT ducing affective reactions and ergonomic mugs intended to be sat- Why do gift-givers often seem to fail at giving what their re- isfaction-inducing. Participants, either as a giver or as a receiver ceivers want? In the present paper, we identify a distinct cause for the depending on condition, first rated how much they liked each option, disparity between givers’ choices and receivers’ satisfaction, which from “just a little (1)” to “very much (7)”. Participants then indicated we term the smile-seeking account: givers primarily aim to induce their preference between the pairs of mugs, on a bipolar scale from desirable spontaneous affective display, such as positive facial ex- strongly preferring the first option (1) to strongly preferring the sec- pression and bodily reactions, as opposed to inducing the greatest ond option (9). receiver satisfaction. Participants then rated the mugs on several measures. We ex- The smile-seeking motive differs from the satisfaction-seeking pected that participants would rate the personalized mugs as induc- motive commonly assumed in the prior literature, which characterizing stronger positive affective reactions than the ergonomic mugs es givers’ primary motive as maximizing receivers’ satisfaction, and (“How much of an affective reaction (e.g., happy facial expressions) treats affective reactions are mere signals of satisfaction, consistent would the receivers show in response to these gifts when receiving with the appraisal theory of emotion (Lazarus 1982). By contrast, them?” on a 7-point Likert scale). Conversely, we expected that the the smile-seeking account builds on a growing literature which sug- more comfortable ergonomic mugs would be rated as yielding higher gests that affective responses can occur without extensive perceptual satisfaction than the personalized mugs (“How much would the re- and cognitive encoding (Forgas 2002, Zajonc 2000). In gift-giving, ceivers be satisfied with these gifts when using them?”) on the same the misalignment between affective reaction and overall satisfaction 7-point scale. gives rise to a discrepancy of motives between givers and receivers: Results givers’ choices may be primarily motivated and gratified by the af- First, the results confirmed that participants expected the per- fective consequences of gift-giving, whereas receivers’ satisfaction sonalized mugs to induce greater affective reactions than the ergo- may be determined primarily by the more deliberative and holistic nomic mugs (M =5.79 vs. 4.59, t(212) = 5.08, p <.001), while cognitive appraisal of the gift, albeit also informed by the initial reaction the mugs were expected to yield similar levels of receiver satisfaction spontaneous affective reactions to the gift. (M 5.31 vs. 5.34, t(212) = -7.70, p < .001). The anticipated We present six studies that test the impact of the smile-seeking satisfaction = reaction benefit, relative to the satisfaction benefit, was significantly account on gift choices and on post-giving enjoyment. We present greater for the option intended to be reaction-inducing (personalized Study 1 with full details on data screening, pretesting, additional mugs ) than for the option intended to be satisfaction-inducing (ergo- measures, and mediation analyses. We then describe with less details nomic mugs; b = +.48, b = -.76, t(212)=8.88, p<.001). Studies 2 and 3A in which same procedures and analyses were used. personalized ergonomic Next, we tested our main hypothesis, comparing givers’ interest Last, we present the main results in Studies 3B, 4A and 4B in which in giving the mugs to receivers’ interest in getting the mugs. The per- data analyses were based on rating and coding of actual gifts. sonalized mugs were more preferred by the givers than the receivers (5.37 vs. 4.81, t(211) = 2.10, p = .037, d = .29), while the ergonomic Study 1: The Giver-Receiver Discrepancy mugs were more preferred by the receivers than the givers (4.15 vs. In Study 1, we test whether givers and receivers differ in their 4.87, t(211) = -2.99, p = .003; ANOVA interaction, F(1,211) = 8.86, relative preference for a more affectively appealing gift in the context p = .003). Likewise, givers’ strong relative preference for the person- of a gift-registry, to hold constant receiver interest in the gifts and alized mugs significantly differed from receivers’ between the two isolate the effect of anticipated affective reactions. mugs on the bipolar scale (3.66 vs. 4.90, t(211) = -2.95, p = .004). We aimed for approximately 100 participants per cell, based on This giver-receiver preference discrepancy was not explained an expected medium effect size (d = .05) and the use of 30-100 per by differences in beliefs about the receivers’ satisfaction with the gift. cell in prior gift-giving research (e.g, Gino and Flynn 2011, Zhang Givers did mispredict receivers’ higher expected satisfaction with and Epley 2012). We recruited 240 adult participants from Amazon the ergonomic mugs (Givers’ predictions of receivers’ satisfaction: MTurk, paid $1 each, yielding 213 completes (M = 33, 52% Male) age M =5.42 vs. M =5.17; Receivers’ expected satisfaction: after excluding invalid participants (duplicate IP addresses or hav- personalized ergonomic M =5.21 vs. M =5.51; interaction F(1,211) = 8.16, p = ing failed an instructional attention check in the end (Oppenheimer, personalized ergonomic .048). However, this misprediction did not explain the preference Meyvis and Davidenko 2009). The exclusion criterion was used in discrepancy between givers and receivers, which persisted control- all studies. ling for expectations of receivers’ satisfaction (β = .682, p =.030). Participants read a scenario about a couple planning their wed- role Instead, the discrepancy was largely explained by anticipated ding, who had sent a wedding registry to a group of friends, each of affective reactions. Givers reported a bigger difference -in receiv whom would choose one of the gifts to give. In the scenario, the ers’ anticipated affective reactions between the two mugs than did couple would receive all the gifts listed on the registry, so the net receivers (Givers: 5.89 vs. 4.10; Receivers: 5.70 vs. 5.07; interac- outcome to the receivers was held constant, regardless of the giv- tion F(1,211) = 23.09, p < .001). Controlling for the difference in er’s choice. The receivers would open the gifts at a wedding shower anticipated affective reactions, which predicts preferences between where all the friends would be present. Participants were randomly the two mugs, we no longer observe a significant giver-receiver dis- assigned to either imagine they were the gift-giver or the gift-receiv- crepancy (β = .219, p > .250; β =.919, p <.001). The effect of er in the scenario. role reaction role (giver vs. receiver) on the relative preference between mug op- Participants saw pictures and descriptions of two similarly tions was fully mediated by differences in expectations of receivers’ priced pairs of mugs, personalized mugs intended to be seen as in-

=16.1, p <.001), but not when they would be absent during gift-re- affective reactions (β = .545, CI = [.277, .813], p <.001), controlling ception (27.7% vs. 26.4%, χ2 = .05, p >.250; interaction (F(1,445) for expected satisfaction. We find similar results using the bipolar =7.78, p =.006). The effect of role on choices in the observable- preference scale. reaction condition held (β = .17, t(224)=3.91, p < .001) when The preference discrepancy also persists when controlling for role controlling for predicted recipient satisfaction. The giver-receiver other factors identified in prior literature as influencing gift-giving, preference discrepancy was again mediated by differences in antici- including hedonic or practical perceptions of the gift, how much pated reactions in the observable-reaction conditions (β = -.058, CI a gift was associated with indulgence or guilt, givers’ and receiv- = [-.101, -.015], p =.008), controlling for anticipated satisfaction, but ers’ construal level (desirable vs. feasible) and regulatory focus no indirect effect was found in the unobservable-reaction conditions (approach-avoidance), as well as social closeness between giver and (β = -.034, CI = [-.088, .020], p = .219). Givers only sought smiles receiver (ps < .025). Response time, gender and age did not affect when they would be there to see the receivers’ smile. gift preferences or moderate the effect of role. In Study 3B, we recruited 198 actual gift-givers (M =33, In sum, givers preferred to give the mug set that receivers liked age 55%Male) online three days before Christmas. Participants each less. This discrepancy was not explained by givers having mistaken specified three gifts they had prepared to give, listed the receivers, beliefs about what receivers would be more satisfied with, the pre- rated both the anticipated affective reaction and expected receiver dominant account of such discrepancies in the prior literature (e.g., satisfaction for each gift, and indicated whether they would be pres- Cavanaugh, Gino and Fitzsimons 2015; Waldfogel 1993, Zhang and ent when each gift would be received. Epley 2009). Instead, the results suggest that the givers’ preferences The gifts participants chose did not differ in the anticipated diverged from what receivers liked better due to the givers’ “smile- level of satisfaction based on whether the givers expected to be pres- seeking” motive: givers chose the option that they believed would ent or not (self-reported M = 5.47 vs. 5.39, b = -.079, p > result in a more desirable affective reaction. satisfaction present .250). In contrast, participants rated their gifts as inducing margin- Study 2: Focus on Reaction vs . Satisfaction in Valentine’s ally stronger affective reactions when they would be giving it in person (M = 6.01 vs. 5.71, d = .27; b = .271, p = .069, observed Day Gifts reaction present power = 90.7%). These findings held when we use independent cod- We recruited 295 relationship partners (M =35, 51% Male) age ers’ ratings instead of participants’ self-report. The effect of presence online to make gift choices for the next day - Valentine’s Day. Af- on how reaction-inducing the gifts were (per coder ratings) also per- ter participants indicated their gender and their partner’s first name, sists controlling for social closeness to the recipient and cost of the male participants were asked to make choices for three pairs of op- gift (b = .184, t=3.15, p = .002). tions, one more reaction-inducing (e.g., blooming roses) and the oth- present er inducing more overall satisfaction (e.g., rose buds) as givers and Study 4A & 4B: Post-Giving Outcomes and Enjoyment female participants evaluated the same gifts as receivers. Participants In Study 4A, we asked 80 participants M = 34, 43%Male) to then predicted the receivers’ affective reactions and satisfaction for Age recall recent favorite and least favorite gifts as either givers or re- each item. Last, we measured personality traits, current relationship ceivers, and rated each gift on reaction and satisfaction scales. Over- status, length of relationship, closeness with partner, and age. all, whether a giver liked or disliked a gift was primarily predicted by We replicated the giver-receiver preference discrepancy for all how reaction-inducing the gift was (β = .199, p < .001, β three pairs of gifts (average choices 52.3% vs. 40.3%, repeated-mea- reaction satisfaction = -.076, p = .133). By contrast, receivers’ liking of gifts related to sures p =.001). As in Study 1, the preference discrepancy between both how reaction-inducing and even more so to how satisfaction- givers and receivers persisted when controlling for predicted satis- inducing the gift was (β = .085, p = .033, β = .215, p < faction (β = .092, p =.007). Moreover, we find the same mediation reaction satisfaction role .001). In sum, the discrepancy between givers’ and receiver’s prefer- via differences in anticipated reaction on preference discrepancy (β ences persists even after the gift reception, which may further rein- = .057, bootstrapped p =.028), even controlling for the indirect effect force the givers’ “smile-seeking” motive in future gift choices. via anticipated satisfaction. In Study 4B, we used a longitudinal design to track givers’ preference and enjoyment after actual gift-giving. We recruited Study 3A & 3B : Discrepancy Eliminated When Receiver 87 Christmas gift-givers before and after Christmas (Mage = 37, Reactions Are Unobserved 58%Male). Participants indicated whether they had been present for Next, we further test whether anticipating being present to ob- the gift reception, rated their perception of the receiver’s immediate serve the receivers’ reaction affects givers’ choices, as predicted by reaction to the gift, their perception of the receiver’s longer-term sat- the “smile-seeking” motive. In Study 3A, we randomly assigned 490 isfaction with the gift, and their own enjoyment from having given mturkers (M =35, 44%Male) to 2(role: giver vs. receiver) x 2(ob- age each gift. servable vs. unobservable) between-subjects conditions to imagine Reaction-inducing gifts resulted in greater giver enjoyment, via buying a blue-tooth speaker as a birthday-gift for a close friend (list- greater receiver reaction that givers reported (indirect effectβ = 2.93, ed by participants themselves). In addition, participants either imag- CI = [.98, 6.58], p <.001), whereas satisfaction-inducing gifts did not ined that the gift-giver would be present on the receiver’s birthday significantly contribute to givers’ enjoyment, even though givers re- and give the gift in person or that the gift-giver would be out of town ported perceiving greater receiver satisfaction for such gifts (indirect and have the gift mailed. The giver was faced with two same-price effect β = .55, CI = [-.66, 2.34], p =.45). Moreover, givers enjoyed upgrade options: gift-wrap (pretested as inducing initial affective the gift-giving experience substantially more if they had been pres- reaction) and a LED-light (pretested as inducing satisfaction dur- ent to see the receivers’ reactions (87.1 vs. 46.3, β = 40.9, p <.001), ing use).The anticipated reaction benefit was greater relative to the even controlling for interpersonal closeness (β = 38.7, p <.001). Con- satisfaction benefit for the option intended to be reaction-inducing sistent with Study 3, observing the receiver’s reaction was critical to (gift wrap) than for the option intended to be satisfaction-inducing the giver’s enjoyment of gift-giving. (t(448)=18.4, p<.001). Givers prefer the gift-warp upgrade (relative to receivers) when they would be present during gift-reception (44.8% vs. 19.8%, χ2 984 / The “Smile-Seeking” Giver: How Immediate Affective Reactions Motivate and Reward Gift-giving General Discussion REFERENCES The forgoing studies offer evidence for a “smile-seeking” mo- Cavanaugh, L. A., Gino, F., & Fitzsimons, G. J. (2015). When tive underlying gift-giving behaviors, challenging previous assump- doing good is bad in gift giving: Mis-predicting appreciation tions that giver-receiver discrepancies are due to givers mispredict- of socially responsible gifts. Organizational Behavior and ing receiver satisfaction. This is an important distinction because Human Decision Processes, 131, 178-189. the receiver’s display of affective reactions to a gift may often differ Forgas, J. P. (2002). Feeling and doing: Affective influences on from the receiver’s satisfaction with the same gift. Facing a tradeoff interpersonal behavior. Psychological inquiry, 13(1), 1-28. between gifts that induce greater reaction and gifts that yield greater Gino, F., & Flynn, F. J. (2011). Give them what they want: satisfaction, givers’ preference for reaction-maximizing gifts will de- The benefits of explicitness in gift exchange.Journal of viate from what their friends, relationship partners and family mem- Experimental Social Psychology,47(5), 915-922. bers would prefer to receive. Furthermore, the givers’ smile-seeking Lazarus, R. S. (1982). Thoughts on the relations between emotion preference was contingent on anticipating an opportunity to observe and cognition. American psychologist, 37(9), 1019. receivers’ affective reactions during gift reception. When givers be- Oppenheimer, D. M., Meyvis, T., & Davidenko, N. (2009). lieve that they would not observe the receiver’s reaction, the giver- Instructional manipulation checks: Detecting satisficing to recipient discrepancy was eliminated. The smile-seeking preference increase statistical power. Journal of Experimental Social also persisted after giving, as givers derived more enjoyment from Psychology, 45(4), 867-872. the receivers’ display of affective reactions. Waldfogel, J. (1993). The deadweight loss of Christmas. The Nonetheless, we do not wish to attribute all underappreciated American Economic Review, 83(5), 1328-1336. gift choices to the smile-seeking motive. Bad gifts can occur due to Zajonc, R. B. (2000). Feeling and thinking: Closing the debate over obvious carelessness (e.g., candies for a diabetic patient), outright the independence of affect. selfishness (e.g., a vacuum cleaner for a romantic partner), or mere Zhang, Y., & Epley, N. (2009). Self-centered social exchange: lack of information. Also different from the smile-seeking motive, differential use of costs versus benefits in prosocial reciprocity. some gift choices may be instead guided by motives to form charac- Journal of personality and social psychology, 97(5), 796. ter or endow assets instead (e.g., parents giving educational materi- Zhang, Y., & Epley, N. (2012). Exaggerated, mispredicted, als or funds to children as birthday gifts). and misplaced: When “it’s the thought that counts” in gift Our findings have implications for research on gift-giving and exchanges. Journal of Experimental Psychology: General, interpersonal decision-making. In particular, our findings call into 141(4), 667. question prior assumptions that people making interpersonal choices primarily intend to maximize receiver satisfaction, and that others’ affective reactions are mere behavioral consequences of their appraisal outcome. The Effect of Pre-giving Incentives on Relationship Norms and Donation Behavior Bingqing (Miranda) Yin, University of Kansas, USA Yexin Jessica Li, University of Kansas, USA Surendra Singh, University of Kansas, USA

EXTENDED ABSTRACT with non-monetary PGIs (p = .08). Monetary PGIs also led partici- A pre-giving incentive (hereafter, PGI) is the provision of a ben- pants to perceive the charity as less communally oriented than no efit/favor before requesting compliance (e.g., Regan 1971). Inthe PGIs (p < .001) and non-monetary PGIs (p < .01). Monetary PGIs in- past three years, PGIs such as coins and greeting cards, were includ- creased the perceived exchange orientation of the charity compared ed in approximately half of the total non-profit mail volume (Parady- with no incentives (p = .02). No difference was found in all other sz + PM Digital Research 2016). Including PGIs to potential donors conditions (all ps > .17). Using PROCESS (model 4; Hayes 2013) requires resources that are already scarce for many nonprofits. Do with 10,000 bootstrap samples, we showed that the relative indirect the benefits of PGIs justify the costs? Additionally, it is unclear how effect of monetary versus no incentives on donation amount was me- different types of PGIs (e.g., coins, mailing labels) influence donor diated by perceived communal orientation of the charity (bootstrap psychology and behavior. Although abundance of research demon- CI: 1.6144, 7.5409) but not by the perceived exchange orientation of strates the persuasive power of reciprocity (Cialdini, 1993), we show the charity (bootstrap CI: -.0415, 3.8874). that in certain pre-giving situations, providing gifts or favors before Similarly, results demonstrated an indirect effect of monetary requesting compliance might have no effect or even backfire. versus non-monetary PGIs on donation amount, also mediated by Due to charities’ focus on communal obligations (McGraw et al. perceived communal orientation (bootstrap CI: .9003, 5.9651) but 2012), we propose that people use communal norms when interact- not exchange orientation (bootstrap CI: -.2086, 2.6474). Overall, the ing with charities, giving benefits (usually time or money) without results support our hypothesis that that lower communal orientation expecting comparable (or any) benefits in return. However, the per- (but not exchange orientation) leads people to donate less. ception of charities as communally oriented can be influenced by the In experiment 3 (N=9,000), we partnered with a local charity type of PGI (monetary vs. non-monetary) enclosed in charity request to examine the effects of monetary ($0.25), non-monetary (a greet- letters. Empirical research suggests that reminders of money dimin- ing card), and no PGIs on donations in the field. Our response rate ish communal motivations (Vohs, Mead, and Goode 2006) and lead (0.52%) was comparable to that of other donor acquisition cam- people to perceive themselves to be in a businesslike or exchange paigns (Smart Annual Giving 2013). For average donation amount relationship with others (Jiang, Chen, and Wyer, 2014). This line of per new donor acquired, monetary PGIs generated less money than research suggests that enclosing low value monetary PGIs evokes no PGIs (Mmonetary =$17.48, Mnon-monetary =$30.45, Mcontrol = $47.50; F (1, exchange norms and leads the charity to be perceived as relatively 45) = 10.36, p < .01). No donation difference was observed between less communally oriented. In addition, donors who follow exchange non-monetary PGIs and no PGIs or between monetary PGIs and non- (vs. communal) norms may reciprocate by giving the charity an monetary PGIs (ps > .1). We also found that, on average, each mail- amount roughly equivalent to what they received. Consequently, we ing with a monetary PGI cost the charity $.59, which is significantly predict that the inclusion of low value monetary PGIs will lead to more than loss per mailing with non-monetary incentives ($.49), F lower donations than when no incentives are provided. Low value (1, 8998) = 11.20, p < .001 or loss per mailing with no PGIs ($.25), non-monetary PGIs, on the other hand, should not evoke exchange F (1, 8998) = 21.65, p< .001. Extrapolated, for our campaign with norms. Thus, donors receiving a non-monetary PGI should use com- 9,000 individuals, including monetary PGIs (versus no incentives) munal norms as they do when they receive no incentive. However, resulted in an additional $.34 net loss per mailing ($1,020 in total) because communal norms dictate non-contingent, need-based giving while including non-monetary PGIs resulted in an additional $.24 net and do not necessitate immediate return of benefits (Aggarwal 2004; loss per mailing ($720 in total). Clark and Mills 1993), a low value non-monetary PGI is unlikely to The present work provides important theoretical and practi- increase donations. cal implications. We integrate research on reciprocity, relationship We tested our predictions in a pilot study, 3 lab experiments, norms, and prosocial behavior and suggest situations where gifts and a field experiment. A pilot study (N=113) found that people per- may not prompt reciprocal behaviors. Specifically, we show that reciprocity after receiving a low value monetary PGI results in lower ceived charities as more communally than exchange oriented (Mcom- donations than comparable value non-monetary or no PGIs under munal = 5.45, Mexchange = 3.38, F(1, 93) = 57.19, p < .001), and busi- exchange norms. This is because external cues such as PGIs affect nesses as more exchange than communally oriented (Mcommunal = 3.48, perceptions of the charity and thus the use of communal versus ex- Mexchange = 4.92, F(1, 93) = 29.18, p < .001). Experiment 1A (N=175) and 1B (N=132) randomly assigned change norms. Practically, our results imply that charitable organiza- participants to view charity letters with a monetary PGI ($.25 in 1A, tions should not be encouraged to use PGIs in soliciting donations. $.50 in 1B), a non-monetary PGI (a greeting card in both studies), or no PGI. Participants reported how much they would like to donate to REFERENCES the charity hypothetically in experiment 1A and from their study pay- Aggarwal, Pankaj (2004), “The effects of brand relationship norms ment ($5 in $1 bills) in experiment 1B. In both studies, people who on consumer attitudes and behavior,” Journal of Consumer received monetary incentives donated less than those who received Research, 31 (June), 87-101. non-monetary PGI or no PGI (ps < .05). No differences emerged be- Cialdini, Robert. B. (1993). Influence: Science and practice, New tween the non-monetary PGI and no PGI conditions (ps > .30). York: HarperCollins. In experiment 2 (N=129), we examine the mediating role of Clark, Margaret S. and Judson Mills (1993), “The Difference communal and exchange orientation on donation behavior. Results between Communal and Exchange Relationships: What It Is showed that monetary PGIs elicited significantly less donations com- and Is Not,” Personality and Social Psychology Bulletin, 19 pared with no PGIs (p = .02) and marginally less donations compared (December), 684–91.

Hayes, Andrew. F (2013). Introduction to Mediation, Moderation, Paradysz + PM Digital Research (2016), “Market Trends 2015 and Conditional Process Analysis: A Regression-Based Non-profit Sector: Direct Mail promotion and Universe Approach. New York, NY: Guilford Press. Trending,” Research report, Paradysz. Jiang, Y., Chen, Z., and Robert S. Wyer Jr. (2014), “Impact of Regan, Dennis T. (1971), “Effects of favor and liking on money on emotion expression,” Journal of Experimental compliance,” Journal of Experimental Social Psychology, 7 Social Psychology, 55 (November), 228–33. (November), 627-39. McGraw, Peter A., Janet A. Schwartz, and Philip E. Tetlock (2012), Vohs, Kathleen D., Nicole L. Mead, and Miranda R. Goode (2006), “From the Commercial to the Communal: Reframing Taboo “The Psychological Consequences of Money,” Science, 314, Trade-Offs in Religious and Pharmaceutical Marketing,” 1154–56. Journal of Consumer Research, 39 (June), 157–73. Choosing versus Rejecting: The Effect of Decision Modes on Subsequent Preferences Sangsuk Yoon, Temple University, USA Vinod Venkatraman, Temple University, USA

EXTENDED ABSTRACT condition may have biased subsequent choice, relative to the reject- We are faced with the same decisions multiple times in many ing condition. circumstances, and often do not employ a single strategy in making these decisions (Venkatraman, Payne, and Huettel 2014; Yoon, Study 3

Vo, and Venkatraman 2017). For instance, one might sometimes pur- In Study 3 (N = 106, Mage = 21.38, SD = 3.97, 66.98% female), chase a product by actively choosing what they like, and other times we used a free-choice paradigm to elucidate how the choosing and by passively rejecting unattractive alternatives (Meloy and Russo rejecting decision modes affected post-choice preferences. In Phase 2004; Shafir 1993). While previous studies have shown that people 1, participants rated attractiveness of a series of 40 items. In Phase focus on compatible attributes (i.e., positive attributes when choos- 2, items were paired based on the similarity of attractiveness ratings ing vs. negative attributes when rejecting), little is known about the from Phase 1, and participants in the choosing condition chose one underlying mechanisms and the impact of these decisions on sub- item they liked, while those in the rejecting condition rejected one sequent preferences. Across three studies, we sought to understand item they disliked. In Phase 3, participants rated the attractiveness how these two decision modes (choosing vs. rejecting) influence sub- ratings of the 40 items again. We were primarily interested in how sequent preferences for the chosen or non-rejected items. the two decision modes affected perceived attractiveness in Phase 3 relative to Phase 1. Consistent with Study 1, participants spent longer

Study 1 in rejecting than in choosing (Mchoosing = 3.13 seconds vs. Mrejecting = 3.70 seconds; t(104) = -2.79, p = .006) in Phase 2. We found that the Seventy-two participants (Mage = 19.46, SD = 1.52, 36% female) were randomly assigned either to choosing or rejecting conditions. gap between preferred items and non-preferred items were greater in The experiment consisted of two phases, and participants made bi- the choosing condition (chosen: M = 0.49, SD = 0.33 vs. non-chosen: nary choice decisions between pairs of items chosen from a set of 48 M = -0.49, SD = 0.33) than in the rejecting condition (non-rejected: triplets in each trial. In Phase 1, two items were randomly selected M = 0.41, SD = 0.32 vs. rejected: M = -0.41, SD = 0.32) in Phase 3. from each triplet and presented. In the choosing condition, partici- The results imply that post-choice preference changes were greater pants chose the item they liked, while in the rejecting condition, they in choosing than in rejecting and the greater post-choice preference rejected the item they disliked. In Phase 2, the chosen/non-rejected changes in the choosing condition may have affected subsequent (target) item from Phase 1 was paired with the remaining item (new) choice more relative to the rejecting condition. from the triplet, and participants chose the item they liked from the pair. We hypothesized that the preference for “target” items will be Conclusion and General Discussion greater for the chosen item from Phase 1 relative to the non-rejected Our results show that choosing increases preference for the cho- item. sen item in a subsequent choice, while rejecting does not. We con- Consistent with previous studies (Chen and Proctor 2017; Soko- jecture from eye-tracking data that participants focus on both relative lova and Krishna 2016), participants in the rejecting condition spent attractiveness and unattractiveness in the rejecting condition provid- longer time than those in the choosing condition in Phase 1 (choosing no advantage to the non-rejected item and no disadvantage to the ing: M = 4.14 seconds, SD = 1.24 vs. rejecting: M = 4.61 seconds, SD rejected item. However, they form positive overall impression for the = 1.86; F = 8.06, p = .006). We found that the two decision modes chosen item in the choosing condition, which subsequently biases asymmetrically influenced subsequent choices. As hypothesized, the preference for the same item in future decisions. This was support- target items were selected more when they were chosen than non- ed by the holistic search pattern from eye-tracking data and greater rejected from Phase 1 (t(70) = 2.37, p = .021). Further analysis re- post-choice preference changes in the choosing condition than in the vealed that choice of the target items in the choosing condition was rejecting condition. We hope to extend these to multi-attribute de- significantly higher than chance level (chance level = .66; t(35) = cisions in future studies to specifically understand if these decision 3.13, p = .004), while that in the rejecting condition was at chance modes lead to increased emphasis on compatible attributes. level (t(35) = -0.82, p > .250). Therefore, choosing increased the attractiveness of chosen item in a subsequent choice, while rejecting REFERENCES did not decrease the attractiveness of the non-rejected item. Chen, Jing and Robert W Proctor (2017), “Role of Accentuation in the Selection/Rejection Task Framing Effect,”Journal of Study 2 Experimental Psychology: General, 146 (4), 543-68. Meloy, Margaret G and J Edward Russo (2004), “Binary Choice In Study 2 (N = 90, Mage = 20.47, SD = 1.86, 56.67% female), we used eye-tracking to elucidate the processes underlying these asym- under Instructions to Select Versus Reject,” Organizational metric preferences using a similar paradigm. We were primarily in- Behavior and Human Decision Processes, 93 (2), 114-28. terested in mechanisms underlying the two decision modes in Phase Shafir, Eldar (1993), “Choosing Versus Rejecting: Why Some 1. We found that participants in the rejecting condition switched their Options Are Both Better and Worse Than Others,” Memory & fixations more between the two items (M = 4.09, SD = 1.25) than cognition, 21 (4), 546-56. those in the choosing condition (M = 3.48, SD = 1.25; t(87) = -2.37, Sokolova, Tatiana and Aradhna Krishna (2016), “Take It or Leave p = .020). Additionally, participants in the choosing condition spent It: How Choosing Versus Rejecting Alternatives Affects longer durations on preferred item than non-preferred item, while Information Processing,” Journal of Consumer Research, 43 those in the rejecting condition spent longer duration on non-pre- (4), 614-35. ferred item than preferred item. We contend that the increased attention to chosen alternative, and the focused search pattern in choosing

Venkatraman, Vinod, John W Payne, and Scott A Huettel (2014), Yoon, Sangsuk, Khoi Vo, and Vinod Venkatraman (2017), “An Overall Probability of Winning Heuristic for Complex “Variability in Decision Strategies across Description‐ Risky Decisions: Choice and Eye Fixation Evidence,” Based and Experience‐Based Decision Making,” Journal of Organizational Behavior and Human Decision Processes, 125 Behavioral Decision Making. (2), 73-87. Don’t Kill the Suspense: The Duality of Suspense in Entertainment Consumption Mujde Yuksel, Suffolk University, USA Elizabeth G Miller, University of Massachusetts Amherst, USA Easwar Iyer, University of Massachusetts Amherst, USA

EXTENDED ABSTRACT they expected to feel during the flow of events and about how the epi- Numerous research studies and anecdotal evidence reveal that sode/game would end. Consistent with our hypotheses, participants sports programs are much more likely to be watched live than oth- were more likely to watch a program when they did not know the out- er types of programs (e.g., Bernoff 2004; Nielsen 2010). Why do come (Mnoknowledge = 6.00, Mknowledge = 4.91 p < .05). This relationship consumers prefer to watch sports events live vs. taped? To answer was mediated by process suspense, but not by outcome suspense (in- this question, we examine the nature of suspense more generally direct effect for process suspense: 95% CI [-3.20, -.33]; and indirect and when outcome knowledge (or spoilers) reduces suspense and effect for outcome suspense: 95% CI [-.95, 1.38]). In other words, consequent consumption enjoyment. While we focus on this ques- outcome knowledge negatively influenced the likelihood of watching tion largely within the context of sports viewership, we note that our the program only when it diminished the process suspense. Further- findings have implications for any experience that is engaged in due more, we found that the relationship between outcome knowledge and to suspense or excitement and which is repeated over time, such as process suspense was moderated by program type such that outcome effectiveness of television advertisements (Elliott 2013), use of re- knowledge diminished the process suspense only for programs with runs, and gift-giving. high correlation between outcome and process suspense (taped sport Drawing on suspense literature (e.g., Duckworth 2012), we ar- event), and not for programs with low correlation (taped TV show; gue that lack of knowledge about an outcome is primarily limited to indirect effect for interaction: 95% CI [.03, .57]). the cognitive processing of information related to estimating the prob- Study 3 (n=56) provides additional evidence that the differ- abilities of various outcomes, including the preferred one. However, ences observed in viewership preferences relates to the relationship a reader/viewer can continue to experience suspense and the resultant between outcome and process suspense, and not just differences in tension via an emotional pathway. That is, one is able to now focus genre. Participants viewed one of four sports scenarios according more on the unfolding of the event and follow it in great detail since to a 2 (outcome-process dependency: low [snowboarding], high [ice one need not invest any resources in predicting the outcome. We refer hockey]) x 2 (outcome knowledge: yes, no) between-subjects design. to these two types of suspense as “outcome suspense” and “process Similar to Study 2, we again find that process suspense (and not out- suspense” respectively. Specifically, outcome suspense refers to anxi- come suspense) mediates the relationship between outcome knowl- ety caused by what an outcome is, while process suspense refers to edge and likelihood to watch (indirect effect for process suspense: anxiety about how that outcome came to be. We propose that spoilers 95% CI [-2.10, -.19]; and indirect effect for outcome suspense: 95% (outcome knowledge) spoil (reduce enjoyment) only in situations in CI [-.19, 1.41]). In addition, the type of program moderated the ef- which outcome suspense and process suspense are highly correlated fect of outcome knowledge on process suspense such that outcome (high outcome-process dependency). Further, we propose that the knowledge had a bigger impact on process suspense for sport events relationship between outcome suspense and process suspense varies with high outcome-process dependency (ice hockey) compared to by program type, and show how process suspense impacts consumer sport events with low outcome-process dependency (snowboarding; preferences for time-shifting TV entertainment consumption. indirect effect for interaction: 95% CI [-2.65, -.42]). Study 1 shows that outcome-process dependency varies with Thus, across three studies, we find that outcome knowledge reduc- program type, and program type influences preferences for real- es enjoyment only in situations in which outcome suspense and process time vs. tape-delayed viewing. Eighty participants were asked how suspense are highly correlated, a situation which characterizes many likely they would be to watch their favorite sport or their favorite sporting events. Further, we show that the impact of spoilers on felt (non-sport) television program at different times (live or after 1 day, suspense helps explain consumer decisions not to time-shift viewing. 2 days, 3 days, 1 week). Consistent with our hypotheses and real- This research contributes to an increased understanding of the world data (cf., Bernoff 2004), participants were significantly more nature of suspense and offers an explanation for the “paradox of suspense” in entertainment consumption. Suspense acts as an important likely to watch sports events broadcast live (Msport = 5.6, Mtv = 4.5, p < .003). They were also willing to pay (WTP) more to watch sports driver of consumer decisions to watch or re-watch shows as well as whether to watch those shows live or time-shifted (i.e., via DVR, on- events broadcast live (WTPsport = $11.09, WTPtv = $6.81), and this WTP drops sharply for sports, but not for TV programs after one day line, etc.). By better understanding the factors that create suspense, marketers can choose more effective marketing messages and can (WTPsport = $1.39, WTPtv = $4.76). Overall, 79% of participants show decreasing WTP patterns for sports compared to 38% for TV (p < better target their entertainment offerings to consumers. .0001). In addition, when asked to explain how watching sports live REFERENCES differed from watching it after its original airing (N=32), 72% of par- Bernoff, Josh (2004), “The Mind of the DVR User: Media and ticipants mentioned the likelihood of learning the outcome if watch- Advertising,” Forrester Research Inc., September 8th. ing the game late and that this knowledge would ruin the experience. Duckworth, AR (2012), “The Paradox of Suspense: I-XII,” The Having confirmed differences in real-time viewing by program Motley View, http://ardfilmjournal.wordpress.com/2012/01/06/ type, studies 2-3 explicitly test our argument that the underlying level the-paradox-of-suspense-i-introduction/ [accessed 5-29-12]. of suspense explains consumers’ likelihood of watching taped events. Elliott, Stuart (2013), “Spoiler alert: Super Bowl ad previews draw Study 2 (n=55) used a 2 (program type: sports event, TV series) x 2 online attention, with criticism,” New York Times, January 30, B1. (outcome knowledge: present, absent) between-subjects experiment. Nielsen (2010), “DVR Use in the US,” online report, http:// Participants imagined knowing or not knowing the outcome of a sport- www.nielsen.com/content/dam/corporate/us/en/newswire/ ing event or television program. They then indicated how likely they uploads/2010/12/DVR-State-of-the-Media-Report.pdf, were to watch the show, as well as how much suspense and uncertainty [accessed 8-8-14].

Advances in Consumer Research 989 Volume 45, ©2017 Hiding Gifts behind the Veil of Vouchers: The Effect of Gift Vouchers in Conditional Promotions Yan Zhang, National University of Singapore, Singapore Yu Ding, Columbia University, USA

EXTENDED ABSTRACT Importantly, this reduced tendency to compare can produce Marketers commonly offer gifts to induce purchases. In the more than just a simple main effect on purchase intention. We suggest marketing literature, such offers are often called conditional promo- that the effect of using a voucher depends on whether the promotion tions, in which consumers receive a supplementary product for free is of high or low value. For high-value promotions, using a voucher or at a discounted price conditional on the purchase of a focal prod- deemphasizes the high value of the promotion and thus decreases uct (Neslin 2002; Palmeira and Srivastava 2013). consumers’ intention to purchase the focal product; whereas for low- Marketers face the question of whether to introduce a voucher value promotions, using a voucher shifts consumers’ attention away when running conditional promotions. For example, a telecommuni- from the low value of the promotion, increasing their intention to cation service provider can offer its users a free pair of headphones purchase the focal product. upon signing up for a package, or hand them a voucher that consum- Our explanation of the effect of vouchers highlights the theo- ers can exchange for the same pair of headphones. As of now, the retical importance of the activation of the comparison process. Our effect of using vouchers as compared to using gifts directly in condi- theory is distinct from a possible alternative that consumers are still tional promotions is a largely under-studied topic. Past research has making the comparisons after a voucher is introduced, but they sim- been mostly devoted to comparing the effect of offering gifts with ply do not apply the comparison outputs to their purchase decisions. that of not doing a promotion at all (Kamins et al. 2009; Palmeira This distinction represents the importance of identifying whether an and Srivastava 2013; Raghubir 2004a; Raghubir 2004b), with that of effect is driven by the activation of a psychological process or by the offering price discounts (Chandran and Morwitz 2006; Krishna et al. application of the outputs of a psychological process (Gilbert and 2002; Liu 2013; Mishra and Mishra 2011; Nunes and Park 2003), or Hixon 1991; Higgins 1996). As of now, no experiments have been with that of offering gifts that are incommensurate with the required conducted to tease apart these two possibilities in the literature on purchases (Kivetz 2005). The current research, instead, compares the intermediaries. effect of directly giving gifts with the effect of using vouchers. In this research, we report experiments that could provide a Consumers’ evaluation of a promotion is driven by the per- theoretical test of the two possibilities. If people make comparisons ceived promotion value, defined as the comparative value of the gift but do not apply the comparison outputs into their decisions, then in relation to the price of the focal product (Value of the gift / Price the effect of vouchers should not be influenced by whether a mental of the focal product) (Krishna et al. 2002). For example, people are comparison process is activated. If, however, vouchers reduce the more likely to try a promotion if it offers a $5 discount for a $15 tendency to compare, then activating a comparison process should calculator than if it offers a $5 discount for a $125 calculator (Kahne- reduce the effect of vouchers. man and Tversky 1984). The comparative nature of the attractiveness The comparison tendency explanation also makes two predic- judgment implies that the evaluation is largely based on the compari- tions that have not been tested previously. First, the extent to which son between the price of the focal product and the value of the gift. using a voucher results in a reduction of comparison tendency can be In this research, we argue that using vouchers reduces price strongly affected by the saliency of the voucher versus the saliency comparison for two major reasons. First, past research has shown of the gift value. We anticipate that if the gift value information is that the acquisition of points and miles can make people focus more made salient even when a voucher is used, the voucher’s effect will on the immediate relationship between invested resources and the diminish. points and miles, and less on the comparative value of the input of Second, the extent to which the promotion value influences pur- resources in relation to the output of reward that really matters (Hsee chase intention also depends on consumers’ familiarity with the focal et al. 2003). Second, research on transaction decoupling finds that product. Research has demonstrated that consumers who are familiar consumers spend more in otherwise identical purchasing situations with the products are more sensitive to the size of price discounts in when paid by credit card than by cash or checks (Feinberg 1986; bundling promotions (Harlam et al. 1995). Thus, we predict that the Hirschman 1979). By adding an intermediate step of swiping one’s effect of vouchers should be more prominent among novices than credit card, the pain of paying is greatly reduced because credit card among experienced consumers. usage effectively dissociates the purchase action (happening on the spot) from the payment action (typically happening once a month) Experiment 1 (Prelec and Loewenstein 1998). Participants were randomly assigned to one of four condi- These findings collectively suggest that adding something in tions in a 2 (promotion value: high vs. low) 2 (voucher: presence between the input and the output impairs people’s ability to see the vs. absence) between-participants design. Participants were told to direct relation between the input and the output. We suggest that the imagine buying a digital camera and were presented with a print introduction of a voucher will produce similar effects. It breaks the advertisement of a camera (including the price information of the direct association between the focal product and the gift (buying the camera). Participants were then informed that if they were to make focal product→receiving a gift) and makes the association indirect the purchase, they would receive a gift. The gift was a 1-TB Toshiba (buying the focal product→receiving a voucher→exchanging for a external hard drive in high-value promotion conditions, and a 4-GB gift). Thus, we anticipate that adding a voucher in between the focal Toshiba flash drive in the low-value promotion conditions. Partici- product and the gift decouples the product from the gift, lowering pants in the voucher-absent conditions were shown pictures of the people’s tendency to assess the comparative ratio between the focal gift directly. Those in the voucher-present conditions were shown a product’s price and the gift’s value. picture of a voucher stating that participants could use it in exchange

Advances in Consumer Research 990 Volume 45, ©2017 Advances in Consumer Research (Volume 45) / 991 for the specified gift. Then participants indicated their likelihood to participants design. We used the same materials from Experiment 1. purchase the digital camera. Besides promotion value and voucher, we also manipulated whether We found a marginally significant main effect of promotion consumers considered the price of the free gift before or after they value, F(1, 157) = 2.76, p = .099, = .017. Participants reported a indicated their purchase intention. We anticipated that making peo- greater intention to purchase the digital camera when presented with ple consider the gift value before indicating their purchase intention a high-value promotion (M = 4.09, SD = 1.39) than low-value pro- would enhance the comparison tendency and this would consequent- motion (M = 3.77, SD = 1.33). Importantly, there was a significant ly act against the effect of a voucher. We also measured comparison interaction between promotion value and voucher, F(1, 157) = 9.09, tendency at the end. p = .003, = .055. For high-value promotions (external hard drive as When participants indicated purchase intention before estimat- gift), using a voucher lowered participants’ purchase intention (M ing gift price, the effect of voucher was replicated,F (1, 315) = 34.66,

= 3.81, SD = 1.33) as compared to presenting the gift directly (M = p < .0001, = .099. The bootstrap analysis revealed that the effect 4.41, SD = 1.40; F(1, 157) = 3.95, p = .049, = .025). In contrast, for of a voucher on purchase intention was mediated by the measured low-value promotions (flash drive as gift), using a voucher increased comparison tendency (ab = -0.11, Boot SE = 0.08, 95% confidence participants’ purchase intention (M = 4.10, SD = 1.28) as compared interval [-0.33 to -0.01]). However, when participants estimated gift to presenting the gift directly (M = 3.42, SD = 1.30), F(1, 157) = price first, the effect of voucher was eliminated, F(1, 314) = 0.04, p 5.19, p = .024, = .032. = .84. The three-way interaction between promotion value, voucher, and evaluation sequence was significant, F(1, 629) = 16.80, p < Experiment 2 .0001, = .026. Experiment 2 replicated the findings of experiment 1 using a design that involved decisions with real consequences. Participants Experiment 5 decided whether to buy a toy (the product). With the purchase of the In real marketing practice, marketers often state the gift price toy, they could also receive two packs of sticky notes for free (the on vouchers. Experiment 5 tested how indicating the gift price on gift). The promotion value was manipulated by changing the price of voucher affects the effect of using vouchers. It used a 2 (promotion the toy. A low toy price makes a high-value promotion and a high toy value: high vs. low) 2 (voucher: presence vs. absence) 2 (gift price price makes a low-value promotion. Participants were either shown information: presence vs. absence) between-participants design. the sticky notes directly (voucher-absence) or a printed voucher that Participants were asked to imagine that they were considering buy- could be exchange for the notes (voucher-presence). Results showed ing a board game (the focal product) for an upcoming party, and that using a voucher in high-value promotions lowered the percent- that a store was running a promotion that buyers of the game would age of participants making the purchase from 63% to 39%, = 3.74, receive either four pieces of Godiva Signature Truffles (high-value p = .053. In contrast, using a voucher in low-value promotions sig- promotion) or one piece of Ferrero Rocher chocolate (low-value nificantly increased the percentage of participants making the pur- promotion). Participants either received a picture of the gift or were chase from 25% to 59%, = 8.87, p = .003. The Logistic Regres- presented with a picture of a voucher. In addition, half participants sion returned a significant interaction between promotion value and received no price information about the gift, the other half were in- voucher, = 11.67, p = .001. formed about the market value of the gift ($10 for Godiva and $0.5 for Ferrero Rocher). We predict that explicitly mentioning the price Experiment 3 of the gift will enhance comparison tendency and thus reduce the Experiment 3 adopted a theoretical method to test the impor- effect of using vouchers. tance of comparison tendency in causing the effect. It used a 2 (pro- The effect of vouchers was replicated when participants were motion value: high vs. low) 2 (voucher: presence vs. absence) 2 not informed of the gift price, with the interaction between promo- (priming: control vs. comparison) between-participants design. We tion value and voucher being significant, F(1, 395) = 8.82, p = .003, primed half participants a comparison mind-set by asking partici- = .022. However, stating gift price mitigated this effect, F(1, 367) = pants to make a series of irrelevant comparisons before presenting 0.94, p = .334, with three-way interaction between promotion value, them the main product and the gift. Participants in the control prim- voucher, and gift price information trending to significant,F (1, 762) ing conditions were asked to make a series of judgments that did not = 1.96, p = .162, = .003. involve comparisons. Then participants then decided whether to buy a toy that coupled with free bookmarks (5 bookmarks in the high- Experiment 6 value condition and 1 bookmark in the low-value condition). Experiment 6 used a 2 (promotion value: high vs. low) 2 We expect to replicate the results of previous experiments in (voucher: presence vs. absence) between-participants design to test the control condition. When people are primed to compare, however, whether the effect of vouchers is moderated by participants’ previous the comparison tendency is enhanced by the priming and the effect experience of purchasing the focal product. Participants imagined of vouchers is predicted to be mitigated. Consistent with this predic- to buy an electronic toothbrush with a gift. The gift was either four tion, we replicated the results in the control conditions with vouchers tubes (high-value promotion) or one tube (low-value promotion) of either increased or decreased purchase intention depending on the toothpaste. Half participants were presented with a picture of the gift, promotion value (interaction =8.57, p=.003), but we didn’t observe and the other half were shown a voucher which could be exchanged similar effects in the conditions where people were primed with a for the gifts. Participants also indicated their previous experience of comparison mindset ( =1.36, p=.244). buying electronic toothbrushes. We split the data by whether one had purchased any electronic toothbrushes before. We predict that those Experiment 4 with past purchase experience were better equipped with price info Experiment 4 adopted a more practically meaningful method to to judge the promotion value, and thus they would be less likely to test the mediating role of comparison tendency. It used a 2 (promo- be influenced by vouchers. tion value: high vs. low) 2 (voucher: presence vs. absence) 2 (se- As expected, the voucher effect was replicated among novices quence: purchase intention first vs. gift price estimate first) between- (F(1, 634) = 45.77, p < .0001, = .067) but not for experienced cus- 992 / Hiding Gifts behind the Veil of Vouchers: The Effect of Gift Vouchers in Conditional Promotions tomers (F(1, 425) = 0.04, p = .85), with three-way interaction be- Hirschman, Elizabeth C (1979), “Differences in Consumer tween promotion value, voucher, and past purchase experience being Purchase Behavior by Credit Card Payment System,” Journal significant,F (1, 630) = 5.63, p = .018, = .009. of Consumer Research, 6 (1), 58-66. Hsee, Christopher K., Fang Yu, Jiao Zhang, and Yan Zhang (2003), General Discussion “Medium Maximization,” Journal of Consumer Research, 30 Our findings further our understanding about how people pro- (1), 1-14. cess information when they are evaluating sales promotions. Al- Kahneman, Daniel and Amos Tversky (1984), “Choices, Values, though consumers usually evaluate a sales promotion by simply as- and Frames,” American Psychologist, 39 (4), 341. sessing its promotion value, our results show that when it is difficult Kamins, Michael A, Valerie S Folkes, and Alexander Fedorikhin to compare the gift value with the focal product price, such as when (2009), “Promotional Bundles and Consumers’ Price a voucher is introduced, consumers’ purchase intention is less influ- Judgments: When the Best Things in Life Are Not Free,” enced by the promotion value. Journal of Consumer Research, 36 (4), 660-70. Our findings also advance research on the effect of- interme Kivetz, Ran (2005), “Promotion Reactance: The Role of Effort- diaries. Besides inducing medium maximization (Hsee et al. 2003) Reward Congruity,” Journal of consumer research, 31 (4), and reducing the pain of paying (Prelec and Loewenstein 1998), our 725-36. research suggests that the decoupling effect created by intermediar- Krishna, Aradhna, Richard Briesch, Donald R Lehmann, and ies could cause changes in how conditional promotions are evalu- Hong Yuan (2002), “A Meta-Analysis of the Impact of Price ated. Although the most effective way of assessing a promotion is to Presentation on Perceived Savings,” Journal of Retailing, 78 directly compare the gift value with the focal product price, people (2), 101-18. seem unlikely to skip the intermediary (in our context, the voucher) Liu, Hsin-Hsien (2013), “How Promotional Frames Affect Upgrade unless they are nudged to do so. Intentions,” Journal of Economic Psychology, 39, 237-48. Our findings suggest that marketers should avoid using vouch- Mishra, Arul and Himanshu Mishra (2011), “The Influence of Price ers for high-value promotions. In comparison, using a voucher that Discount Versus Bonus Pack on the Preference for Virtue and can be exchanged for a relatively inexpensive gift in conditional pro- Vice Foods,” Journal of Marketing Research, 48 (1), 196-206. motions could boost sales while minimizing promotion costs. Neslin, Scott A. (2002), Sales Promotion. Cambridge, MA: Marketing Science Institute. REFERENCES Nunes, Joseph C and C Whan Park (2003), “Incommensurate Chandran, Sucharita and Vicki G Morwitz (2006), “The Price Resources: Not Just More of the Same,” Journal of Marketing of “Free”-Dom: Consumer Sensitivity to Promotions with Research, 40 (1), 26-38. Negative Contextual Influences,”Journal of Consumer Palmeira, Mauricio M and Joydeep Srivastava (2013), “Free Offer≠ Research, 33 (3), 384-92. Cheap Product: A Selective Accessibility Account on the Feinberg, Richard A (1986), “Credit Cards as Spending Facilitating Valuation of Free Offers,”Journal of Consumer Research, 40 Stimuli: A Conditioning Interpretation,” Journal of consumer (4), 644-56. research, 13 (3), 348-56. Prelec, Drazen and George Loewenstein (1998), “The Red and the Gilbert, Daniel T and J Gregory Hixon (1991), “The Trouble of Black: Mental Accounting of Savings and Debt,” Marketing Thinking: Activation and Application of Stereotypic Beliefs,” Science, 17 (1), 4-28. Journal of Personality and Social Psychology, 60 (4), 509. Raghubir, Priya (2004a), “Coupons in Context: Discounting Prices Harlam, Bari A, Aradhna Krishna, Donald R Lehmann, and Carl or Decreasing Profits?,” Journal of retailing, 80 (1), 1-12. Mela (1995), “Impact of Bundle Type, Price Framing and --- (2004b), “Free Gift with Purchase: Promoting or Discounting Familiarity on Purchase Intention for the Bundle,” journal of the Brand?,” Journal of Consumer Psychology, 14 (1), 181-86. Business Research, 33 (1), 57-66. Higgins, Edward Tory (1996), “Knowledge Activation: Accessibility, Applicability, and Salience,” in Social Psychology: Handbook of Basic Principles, E. T. Higgins and A. W. Kruglanski, eds. New York, NY: Guilford Press. Counterhedonic Food Consumption: How Eating “Yucky but Healthy” Food Promotes Marketplace Morality Zhe Zhang, University of Houston, USA Vanessa M. Patrick, University of Houston, USA

EXTENDED ABSTRACT We present a set of four studies to support our theorizing (Fig- With rising awareness of widespread obesity and food-related ure 1). In study 1, we provide evidence to support our main hypoth- health concerns, there has been an increase in the consumption of esis that counterhedonic consumption leads to greater marketplace “yucky, but healthy” foods. This category of food includes items like morality. As predicted, participants perceive counterhedonic food cod liver oil, wheatgrass, kale, and, Marmite, but also foods like ed- (wheatgrass juice, Mcounterhedonic = 4.12) as to be significantly more ible worms, ants, crickets and grasshoppers. Since the consumption disgusting (disgusting, unpleasant, revolting, tasty (reverse coded), of these “yucky, but healthy” foods cannot be adequately explained 1 = not at all, 7 = very much; α = .89) than baseline homeostat- by the fundamental homeostatic (to maintain one’s energy balance) ic food (purified water; Mhomeostatic = 1.99; F (1, 115) = 73.14, p < . or hedonic (reward-based) eating pathways (Lutter & Nestler, 2009), 05). However, participants in the counterhedonic condition showed we herein posit that this category of food consumption is driven by heightened post-consumption morality: they were willing to donate a counterhedonic pathway that complements the homeostatic and more money to charity (M = $13.28) compared to participants in the hedonic pathways already established in food consumption theory. homeostatic condition (M = $8.63; F (1, 115) = 4.85, p < .05). To be categorized as “counterhedonic”, a food has to be “yucky, but Study 2 compares the counterhedonic food pathway with the healthy”; i.e. the food has to be unpleasant in taste/texture/flavor and higher order hedonic food pathway and provides initial support for generate an offensive oral experience that leads to feelings of disgust the underlying mechanism. Results show that even counterhedonic during the consumption process, while delivering significant health food (Marmite spread; Mcounterhedonic = 4.70) was considered signifi- benefits (for example, edible worms, wheatgrass juice, etc.). cantly more disgusting than hedonic food (chocolate spread; Mhedonic We further theorize that since counterhedonic food consump- = 1.97; F (1, 84) = 62.70, p < .05), participants in the counterhedonic tion entails one’s ability to withstand disgusting food in order to condition felt significantly heighted moral self-regard compared to gain its health benefits, the retrospective evaluation of this con- participants in the hedonic condition (Mcounterhedonic = 3.50, Mhedonic = sumption leads consumers to experience heightened moral self- 2.82; F (1, 84) = 5.00, p < .05). In terms of marketplace morality regard post-consumption that translates into marketplace moral- (helping experimenter pick up pencils), participants in the counterhe- ity. Moral self-regard is considered as a reflective view of oneself donic condition were more likely to help the experimenter than those or a moment-to-moment self-concept that answer to the question in the hedonic condition (63% vs. 41%, χ2(1) = 4.05, p < .05), and

“How moral am I” (Monin & Jordan, 2009). Drawing on opponent they also picked up more pencils (Mcounterhedonic = 6.39, Mhedonic = 3.38; process theory (Solomon, 1980) and the retrospective rosy view F (1, 84) = 7.28, p < .05). Mediation analysis indicates significant theories (Ross & Wilson, 2002; Wilson & Ross, 2001), we theorize mediation by post-consumption moral self-regard on consumption that the primary a process for counterhedonic food consumption type and marketplace morality (95% CI = [.0168, .5441]. is aroused at the time of consumption by the sensory properties of Studies 3 and 4 examine the parameters of counterhedonic unappetizing foods, which elicits feelings of disgust or oral offen- consumption as well as identify boundary conditions for the effect. siveness. Such an affective state, however, is short-lived and ter- Specifically, study 3 shows that changing the costliness of the ini- minates when the stimulus is removed (when food consumption is tial behavior by modifying a counterhedonic food (original edible completed). Shortly after, the opponent loop b process is activated worms) to look and sound tastier (chocolate covered worms) dimin- by self-reflection or self-evaluation wherein overcoming the bad ishes both the post-consumption moral self-regard (M original= 4.01, taste and eating something good for one’s own health is considered Mchocolate = 3.13; F (1, 153) = 10.53, p < .05) and marketplace morality positive, virtuous and morally good. (time to volunteer for a charitable event; M original = 10.53 hours, M

Since the desire for a positive moral self-regard motivates chocolate= 7.28 hours; F (1, 153) = 7.38, p < .05). Study 4 replicates and people to engage in prosocial actions (Schaumberg & Wiltermutch, builds upon the findings of study 3 by illustrating that modifying the 2014), and moral affective experiences serve as the energy that drive counterhedonic food to make it more hedonic or by concealing the people to do good and avoid doing bad (Kroll & Egan, 2004), we health benefits of the food similarly diminish moral self-regard and further argue that the positive moral self-regard post-consumption decrease marketplace morality. would orient consumers to a greater good (Keltner, 2016) and can Taken together, these studies demonstrate that counterhedonic translate to enhanced marketplace morality (Kirmani, 2015; Loureiro food consumption is uniquely characterized by the sensory feeling of et al., 2016). core disgust during the consumption of a food that delivers signifi- This research aims to make three theoretical contributions. First, cant health benefits. Post-consumption, these feelings of disgust are we propose a novel food consumption pathway – the counterhedonic replaced by positive moral self-regard, which orients consumers to pathway – to complement and expand upon the already established a greater good and can translate to enhanced marketplace morality. homeostatic and hedonic food consumption pathways. Second, we identify a novel downstream consequence of counterhedonic food REFERENCES consumption – the feeling of positive moral self-regard which in turn Kirmani, A. (2015). Marketplace morality. in Diehl, Kristin and orients consumers to a greater good and promotes prosocial behavior Carolyn Yoon, eds, and marketplace morality. Third, this work lends further support to Advances in Consumer Research, vol. 43. ACR Presidential the notion that food consumption can serve as an antecedent to mar- Address ketplace morality - when we eat well (physical consumption), we can Kroll, J., & Egan, E. (2004). Psychiatry, moral worry, and the moral also feel virtuous and do good (impact on our psychological states). emotions. Journal of Psychiatric Practice, 10(6), 352.

Loureiro, Y. K., Bayuk, J., Tignor, S. M., Nenkov, G. Y., Baskentli, Ross, M., & Wilson, A. E. (2002). It feels like yesterday: Self- S., & Webb, D. (2016). The case for moral consumption: esteem, valence of personal past experiences, and judgments Examining and expanding the domain of moral behavior to of subjective distance. Journal of Personality and Social promote individual and collective well-being. Journal of Psychology, 82(5), 792–803. Public Policy & Marketing, 35(2), 305–322. Schaumberg, R. L., & Wiltermuth S. S. (2014). Desire for a positive Lutter, M., & Nestler, E. J. (2009). Homeostatic and hedonic moral self-regard exacerbates escalation of commitment to signals interact in the regulation of food intake. The Journal of initiatives with prosocial aims. Organizational Behavior and Nutrition, 139(3), 629–632. Human Decision Processes, 123(2), 110-123. Monin, B., & Jordan A. H. (2009). Dynamic moral identity: A Solomon, R. L. (1980). The opponent-process theory of acquired social psychological perspective. In Personality, Identity, and motivation: The costs of pleasure and the benefits of pain. Character: Exploration in Moral Psychology, University American Psychologist, 35(8), 691–712. Press, Cambridge, UK. Wilson, A. E., & Ross, M. (2001). From chump to champ: People’s Oatley, K., Keltner, D., & Jenkins, J. M. (2006). Understanding appraisals of their earlier and present selves. Journal of emotions (2nd ed.). Oxford: Blackwell Publishing. Personality and Social Psychology, 80(4), 572–584. Going It Alone or Together: The Role of Product Space on Consumer Perceptions of Price Promotions Yuli Zhang, Drexel University, USA Hyokjin Kwak, Drexel University, USA Marina Puzakova, Lehigh University, USA

EXTENDED ABSTRACT In Studies 1 and 2, participants were presented with a shelf that Research has recently begun to illuminate the impact of an im- displayed products either with a space or no space between them. portant retail atmospheric variable, namely, the allocation of shelf A price promotion was manipulated by a price tag (e.g., “Now on space to products. For instance, additional space allocated to a prod- sale, $2.50” (single unit) vs. “Now on sale, 2/$5.00” (multiple unit)). uct in a retail setting increases product sales (Dreze et al. 1994) and As predicted, participants in the space condition were less likely to facilitates greater product preferences (Sevilla and Townsend 2016). purchase a product when they viewed a multiple vs. single unit price

However, what is still missing from this literature is the mere ef- promotion (MUPP vs. SUPP) tag (Study1: Mmultiple = 3.24 vs. Msingle fect of the existence (vs. nonexistence) of space between products = 4.35; F(1, 97) = 11.51, p < .01; Study 2: Mmultiple = 4.52 vs. Msingle = on a retail shelf. Space between products is a crucial retail factor to 5.47; F(1, 99) = 4.22, p < .05). In contrast, the MUPP vs. SUPP led to consider because the perceptual representations of space influence greater purchase intentions in the no space condition (Study 1: Mmul- people’s attitudes, thoughts, and judgments (Williams and Bargh tiple = 4.28 vs. Msingle = 3.49; F(1, 97) = 5.63, p < .05; Study 2: Mmultiple

2008). In this research, we put forward a novel idea that space be- = 5.47 vs. Msingle = 4.05; F(1, 99) = 9.29, p < .01). Studies 1 and 2 tween products can impact the effectiveness of a multiple (e.g., two find support for the positive (negative) effect of MUPP vs. SUPP on for $2) versus single (e.g., one for $1) unit price promotion strategy. purchase intentions in the no space (space) condition. Study 2 fur- We build our conceptual framework based on grounded cogni- ther sheds light on the mediating effect of consumers’ distinctiveness tion and sociocultural messages that suggest that people are likely to motivations (consumers’ thoughts about others) on the relationship learn to associate spatial distance concepts with relational distance between the type of price promotions and purchase intentions in the (Meier et al. 2012). Furthermore, previous research posits that the space (no space) condition. activation of the concept of relational closeness is likely to trigger Study 3 examines the moderating effect of product variety in thoughts involving these close others (Aron et al. 1992), whereas the space context. As predicted, participants presented with no prod- activating relational distance is likely to facilitate people’s desires to uct variety indicated lower purchase intentions when they were ex- be different from others (Mashek et al. 2011). Hence, we propose that posed to MUPP vs. SUPP (Mmultiple = 3.89 vs. Msingle = 5.30; F(1,102) the nonexistence of a space between products is likely to activate the = 10.33, p < .01). In contrast, we found no significant differences in concept of relational closeness which in turn leads to thinking about purchase intentions between MUPP and SUPP (p > .70) in the prod- close others, whereas the existence of a space between products is uct variety condition. Further regression analysis results revealed a likely to develop a sense of relational distance that in turn triggers significant indirect path from product variety to purchase intentions consumers’ distinctiveness motivations. through CNFU in the MUPP condition (a point estimate for the effect Prior research also suggests that the quantity anchor in a multi- = -.20; 95% CI = [-.48,-.04]). ple (vs. single) price promotion motivates thoughts about large prod- Study 4 investigates the moderating effect of product popularity uct consumption with others (Manning and Sprott 2007). As such, in the no space context. Participants in the popular product condition we propose that the nonexistence of a space between products, which had greater purchase intentions when they were exposed to MUPP enhances consumers’ thinking about others, is likely to facilitate this vs. SUPP (Mmultiple = 5.25 vs. Msingle = 3.83; F(1,112) = 8.22, p < .01). type of thoughts and, thus, enhance the effectiveness of multiple ver- In contrast, we found no significant differences between MUPP and sus single unit price promotions. In contrast, research also indicates SUPP (p > .90) for participants in the less popular condition. Further that the quantity anchor in a multiple (vs. single) unit price promo- regression analysis results revealed a significant indirect path from tion motivates consumers to buy multiple homogenous items (Man- product popularity to purchase intentions through the thoughts about ning and Sprott 2007). It is logical to assume that consumers with others index in the MUPP condition (a point estimate for the effect = distinctiveness motivations will be less attracted to the idea of con- .35; 95% CI = [.03,.88]). suming multiple homogeneous products. Overall, we propose that Overall, our research is the first to establish the novel relation- when there is a space between products on a retail shelf, multiple (vs. ships between space and multiple versus single unit price promo- single) unit price promotions would be less effective and would lead tions. This work also contributes to prior research on spatial cues by to lower purchase intentions. shedding new light on processes underlying the impact of space on Furthermore, we theorize that as product variety is more likely price promotions, as well as on important boundary conditions. to satisfy consumers’ desires to be different (Kim and Drolet 2003), anchoring on multiple consumption (induced by multiple unit price REFERENCES promotions) of a variety of products should attenuate the negative Aron, Arthur, Elaine N. Aron, and Danny Smollan (1992), effect of multiple (vs. single) unit price promotion on purchase inten- “Inclusion of Other in the Self Scale and the Structure of tions in the space context. We also theorize that because information Interpersonal Closeness,” Journal of Personality and Social regarding lower product popularity is likely to reduce consumers’ Psychology, 63 (4), 596-612. thoughts about consuming these less popular products with others Dreze, Xavier, Stephen J. Hoch, and Mary E. Purk (1994), “Shelf (Zhu and Zhang 2010), anchoring on multiple consumption (induced Management and Space Elasticity,” Journal of Retailing, 70 by multiple unit price promotions) of less popular products should (4), 301-26. attenuate the positive effect of multiple (vs. single) unit price promo- Kim, Heejung S. and Aimee Drolet (2003), “Choice and self- tion on purchase intentions in the no space context. expression: a cultural analysis of variety-seeking,” J Pers Soc Psychol, 85 (2), 373-82.

Manning, Kenneth C. and David E. Sprott (2007), “Multiple Sevilla, Julio and Claudia Townsend (2016), “The Space-to Product unit price promotions and their effects on quantity purchase Ratio Effect: How Interstitial Space Influences Product intentions,” Journal of Retailing, 83 (4), 411-21. Aesthetic Appeal, Store Perceptions and Product Preference,” Mashek, Debra, Benjamin Le, Ken Israel, and Arthur Aron (2011), Journal of Marketing Research. “Wanting less closeness in romantic relationships,” Basic and Williams, Lawrence E. and John A. Bargh (2008), “Keeping one’s Applied Social Psychology, 33 (4), 333-45. distance - The influence of spatial distance cues on affect and Meier, Brian P., Simone Schnall, Norbert Schwarz, and John A. evaluation,” Psychological Science, 19 (3), 302-08. Bargh (2012), “Embodiment in Social Psychology,” Topics in Zhu, Feng and Xiaoquan Zhang (2010), “Impact of Online Cognitive Science, 4 (4), 705-16. Consumer Reviews on Sales: The Moderating Role of Product and Consumer Characteristics,” Journal of Marketing, 74 (2), 133-48. Unpacking Mixed Emotions at Experiential Endings: Total Emotionality, and Differential Roles of Enjoyableness, Meaningfulness and Sociality of Experience Xiaohua Zhao, Tsinghua University, China Wendy Liu, University of California, San Diego, USA

EXTENDED ABSTRACT the control condition, there is neither recalling task nor reading task. When you realize that a positive event is coming to an end, how As the measure of the mixed emotion, we use the 19-item PANAS does the salience of its ending make you feel? This research provides (Watson et al., 1988) and get the mixed emotion by MIN (happiness, a deeper understanding of this emotional state and its antecedents, sadness) (following Hershfield et al., 2008). We found that when and some implications for maximizing consumer experiential and highly enjoyable experience was recalled, people generate higher emotional well-being. mixed emotion in the ending condition than in the no-end condi-

Emotions research shows that people can sometimes experi- tion (Mending= 1.76 vs. Mcontinuing=1.25, p=0.000). But when the mildly ence complex emotional states where positive and negative emotions enjoyable experience was recalled, people just generate marginally concur simultaneously (Williams and Aaker, 2002) and awareness of higher mixed emotion in the ending condition than in the no-end limited time will increase occurrence of mixed emotions (Carstensen, condition (Mending= 1.64 vs. Mcontinuing =1.39, p=0.071). And both in 2006). However, although previous literature focused on mixed emo- highly and mildly enjoyable experience, mixed emotion between the tions produced by endings of meaningful experiences (Routledge et continuing condition and control condition has no significant differ- al., 2011; Baumeister et al., 2013), the role of meaningfulness of the ence. These results support H1 that the more intensely enjoyable the experience in this relationship remains unclear. Specially, it is un- experience, the greater the increase in mixed emotions towards its clear whether the complex emotion of happiness mixed with sadness ending. Besides, planned contrast analysis showed that emotionality is unique to the ending of meaningful experiences, or if complex is significantly lower in the ending condition than in the- continu emotions can generalize to the ending of any positive, enjoyable ex- ing condition (Mending= 3.96 vs. Mcontinuing =4.48, p=0.000), support- periences, even when they are relatively low in meaning. Further, ing H5a that experiential endings subdue rather than heighten one’s previous research tended to examine meaningful experiences inter- overall emotional state (in addition to making it more mixed) than twined with sociality (Hershfield et al., 2008), while in many situ- when the event is continuing. ations meaningful experiences may not be shared with others (i.e., Next, we tested the role of meaningfulness in study 2 and role experiences done alone, Ratner and Hamilton, 2015). Consequently, of sociality in study 3. The manipulation of the meaningfulness and another question that arises is whether mixed emotions at endings sociality is similar to study 1, which used the experience recalling are also determined by the sociality of the experience. Thus one main task. The ending manipulation and dependent measures were the objective of this research is to understand the emotional state of im- same as in study 1. We separated the different roles of enjoyableness pending endings, and in particular, whether enjoyableness, meaning- and meaningfulness in study 2 and employed a 2 (high vs. low enjoy- fulness, and sociality, are critical drivers of the mixed emotion when ableness) × 2 (meaningful vs. meaningless) × 2 (ending vs. continu- the experience approaches an end. We proposed that mixed emotions ing) between-subjects design. A three-way ANOVA showed a main will occur when an enjoyable experience is approaching an end (H1), effect of ending (p=0.000) and a two-way interaction effect of enjoy- higher (vs. lower) level of enjoyableness increases mixed emotions ableness and ending (p=0.006), but no two-way interaction effect of when the experience is coming to an end, independent of meaning- meaningfulness and ending (p=0.076) nor three-way interaction ef- fulness (H2), higher (vs. lower) level of meaningfulness increases fect of enjoyableness, meaningfulness and ending (p=0.702). And the mixed emotions when the experience is coming to an end, indepen- planned contrast results also supported H1, H2, but not H3, which dent of enjoyableness (H3), and sociality per se is not a driver of means that the level of enjoyableness, but not the level of meaning- mixed emotions when facing an end of the experience (H4). fulness of an event critically determines whether one will feel mixed Another question that has not been explicitly explored in previ- emotions when the event is coming to an end. Then we tested H4 ous research is the total amount of emotionality when an experience in study 3, with a 2 (meaningful vs. meaningless) × 2 (social vs. is coming to an end. It may be heightened because of the intensified non-social) × 2 (ending vs. continuing) between-subjects design, and positive and negative emotions induced by endings (Loewenstein, as predicted, we found sociality did not impact how much mixed 1987), and may also be dampened because of the decreased positive emotions were engendered by an approaching ending. Besides, the emotions and increased sadness resulted from endings. These two emotionality analysis all supported H5a. possibilities paint very different pictures for the emotional quality Generally, if marketers wish to maximize total emotionality, of an experiential ending, one intense (H5), and one subdued (H5a), they may not wish to make salient the ending of a positive event; on and this research seeks to uncover which state is the more accurate the other hand, if marketers wish to maximize mixed emotions, mak- description of the affective state during endings. To test our hypoth- ing the end salient is an effective way. In future research, we wish to eses, we conducted 3 studies and the results provide a deeper and examine negative experiences, to see whether it mirrors or reverses, more thorough understanding of the emotional state at experiential and examine the downstream consequences. endings. In study 1, we randomly assigned 456 participants, recruited REFERENCES through MTurk, into five conditions, which is a 2 (high vs. low en- Baumeister, R. F. (1991), Meanings of life, New York, NY: Guilford joyableness) × 2 (ending vs. continuing) + Control between-subjects Press. design. For the enjoyableness manipulation, we asked participants Baumeister, R. F., Vohs, K. D., Aaker, J. L., and Garbinsky, E. N. to recall an experience which is highly enjoyable versus mildly en- (2013), “Some key differences between a happy life and a joyable. For the ending manipulation, they will read a paragraph in- meaningful life,” The Journal of Positive Psychology, 8(6), formed them there is an approaching ending of that experience ver- 505-516. sus no information about ending and they continue to enjoy that. In Becker, E. (1973), The denial of death, New York, NY: Free Press.

Advances in Consumer Research 997 Volume 45, ©2017 998/ Unpacking Mixed Emotions at Experiential Endings: Total Emotionality, and Differential Roles of Enjoyableness, Meaningfulness and Sociality of Experience

Bhattacharjee, A., and Mogilner, C. (2014), “Happiness from Larsen, J. T., and McGraw, A. P. (2011), “Further evidence ordinary and extraordinary experiences,” Journal of Consumer for mixed emotions,” Journal of personality and social Research, 41(1), 1-17. psychology, 100(6), 1095. Carstensen, L. L. (2006), “The influence of a sense of time on Loewenstein, G. (1987), “Anticipation and the valuation of delayed human development,” Science, 312(5782), 1913-1915. consumption,” The Economic Journal, 97(387), 666-684. Ersner-Hershfield, H., Mikels, J. A., Sullivan, S. J., and Carstensen, Schimmack, U. (2001), “Pleasure, displeasure, and mixed feelings? L. L. (2008), “Poignancy: mixed emotional experience in the Are semantic opposites mutually exclusive?” Cognition and face of meaningful endings,” Journal of personality and social Emotion, 15, 81–97. psychology, 94(1), 158. Tiedens, L. Z., and Linton, S. (2001), “Judgment under emotional Ersner-Hershfield, H., Carvel, D. S., and Isaacowitz, D. M. certainty and uncertainty: the effects of specific emotions on (2009), “Feeling happy and sad, but only seeing the positive: information processing,” Journal of personality and social Poignancy and the positivity effect in attention,”Motivation psychology, 81(6), 973. and Emotion, 33(4), 333-342. Ratner, R. K., and Hamilton, R. W. (2015), “Inhibited from bowling Fong, C. T., and Tiedens, L. Z. (2002), “Dueling experiences and alone,” Journal of Consumer Research, 42(2), 266-283. dual ambivalences: Emotional and motivational ambivalence Routledge, C., Arndt, J., Wildschut, T., Sedikides, C., Hart, C. M., of women in high status positions,” Motivation and Emotion, Juhl, J., and Schlotz, W. (2011), “The past makes the present 26(1), 105-121. meaningful: nostalgia as an existential resource,” Journal of Huta, V., and Ryan, R. M. (2010), “Pursuing pleasure or virtue? The personality and social psychology, 101(3), 638. differential and overlapping well-being benefits of hedonic Vallacher, R. R., and Wegner, D. M. (1985), A theory of action and eudaimonic motives,” Journal of Happiness Studies, 11, identification, Hillsdale, NJ: Erlbaum. 735–762. Vallacher, R. R., and Wegner, D. M. (1987), “What do people think Larsen, J. T., McGraw, A. P., Mellers, B. A., and Cacioppo, J. T. they’re doing: Action identification and human behavior,” (2004), “The agony of victory and thrill of defeat: Mixed Psychological Review, 94, 3–15. emotional reactions to disappointing wins and relieving Williams, P., and Aaker, J. L. (2002), “Can mixed emotions losses,” Psychological Science, 15(5), 325–330. peacefully coexist?” Journal of Consumer Research, 28(4), Larsen, J. T., McGraw, A. P., and Cacioppo, J. T. (2001), “Can 636-649. people feel happy and sad at the same time?” Journal of personality and social psychology, 81(4), 684. Exercise Your Mind – Physical Activity Alters Attribute Weighing in Consumer Choice Laura Zimmermann, London School of Economics and Political Science, UK Amitav Chakravarti, London School of Economics and Political Science, UK

EXTENDED ABSTRACT findings indicate that PA leads consumers to focus less overly on Across five studies we find that physical activity (PA) leads to desirability and consider feasibility criteria more in choices that re- improved decision-making in unrelated decision domains. Specifi- quire trade-offs. cally, we find that PA (both measured and manipulated) leads con- In Study 1 we tested exercisers before vs. after visiting a gym sumers to weigh different pieces of information more appropriately (N=90) in a consumer choice task that required making trade-offs be- and improves reliance on relevant product information. The results tween feasibility and desirability attributes (adapted from Liu, 2008). were robust to the inclusion of several control variables in the analy- Participants were asked to imagine they wanted to go on a hiking trip sis. and were presented two options. One of them was characterised by PA has tremendous health benefits as decades of medical and high desirability (scenery with creeks and waterfalls) and low feasi- epidemiological research show (e.g., Moore, et al., 2016; Morris, bility (limited parking, 70 miles away). The other option was char- Heady, Raffle, Roberts, & Parks, 1953). In addition, researchers have acterised by low desirability (scenery with boulders and bushes) and started to investigate the complex relationship between PA, physi- high feasibility (plenty of parking, 40 miles away). Participants were cal wellbeing and brain health. In particular, beneficial effects of asked to choose between both options and to rate how much they PA have been found for emotional health, memory and executive focused on the desirability attribute and the feasibility attribute. Re- functions (Hopkins, Davis, Vantieghem, Whalen, & Bucci, 2012). sults showed that participants considered desirability and feasibility Overall, studies point to the direction that PA enhances cognitive attributes more equally after exercising compared to before (F(1, 88) functions and protects against the development of neurodegenerative = 4.64, p = .034). The results were robust to the inclusion of several diseases (Kramer & Erickson, 2007). Despite the abundant evidence control variables in the analysis. of PA benefits for cognition, little research has investigated the ef- Study 2 (N=257, online) compared participants who engage in fect of PA on judgment and decision-making. We hypothesized that regular PA to non-active individuals. Non-active individuals rated both regular and single bouts of PA would lead to spillover effects on their likelihood to purchase a high desirability/low feasibility option unrelated judgments and decision-making due to improved informa- significantly higher than their likelihood to purchase a low desirabil- tion integration. In particular, we predict that decision makers will be ity/high feasibility option (MHDLF = 61.00 vs. MLDHF = 44.79). Regular better able to ignore unimportant features and not neglect important PA individuals on the other hand showed no difference in their likeli- features of their decision environment. hood to purchase (MHDLF = 53.11 vs. MLDHF = 52.89, F(1, 253) = 4.24, Our conjecture is based on very recent neuroscience research. p = .041). Raichlen, et al. (2016) that argues that PA, such as running, represents Study 3 (N=59, parkrun UK) shows that regular runners with a complex activity for the brain. It involves sophisticated, simultane- significantly better average 5 kilometre running performances tended ous processing and monitoring of internal and external information. to choose a low desirability/high feasibility option rather than a high According to them, areas of the brain related to cognitive functions desirability/low feasibility option (χ2(1) = 4.87, p = .027, β = -.134, such as planning, inhibition, monitoring, attentional switching, and p = .042). Overall, these findings indicate that PA leads decision- multi-tasking are activated when engaging in PA. Nevertheless it makers to focus less overly on desirability and consider feasibility remains unclear whether PA also influences how people make deci- criteria more when making choices that require trade-offs. sions in real-life domains that are unrelated to exercising, and how In studies 4 and 5 (total N=527, online), we find that regular they integrate different pieces of information to make such unrelated PA seems to improve a decision-maker’s ability to rely on relevant decisions. vs. irrelevant information. When faced with irrelevant product infor- In particular, we investigated spillover effects of PA on two well- mation, consumers find it difficult to ignore irrelevant information, researched consumer decision paradigms – the desirability / feasibil- and typically “dilute” their judgments (i.e., their product judgments ity choice conflict (Liberman & Trope, 1998) and the dilution effect are lower). Two studies reveal that PA aids people’s ability to focus (Meyvis & Janiszewski, 2002; Nisbett, Zukier, & Lemley, 1981). If on relevant information and ignore irrelevant information in prod- PA improves the integration of different pieces of information, we uct judgments. When faced with irrelevant information in addition would expect people who engage in PA to apply more appropriate to relevant information, sedentary subjects significantly diluted (i.e., weights to different decision attributes. In the feasibility-desirability lowered) their product judgments. However, there was no significant choice conflict it is commonly observed that decision-makers overly dilution effect for PA individuals F( (1, 283) = 4.56, p = .033). focus on the desirability features and neglect feasibility features; we Across these sets of studies, it appears that PA leads decision- expect PA would lead to less or no neglect of feasibility attributes in makers to weigh different pieces of information more appropriately. product choices, which require trade-offs between desirability and The results remain unchanged after controlling for various confound- feasibility attributes. Conversely, in the dilution paradigm (wherein ing variables including demographics (e.g., education and income), people are unable to ignore irrelevant information and end up “dilut- affect and personality traits (e.g., conscientiousness). Our findings ing” their judgments) this would lead to smaller or no dilution ef- have important implications since they extend the benefits of PA to fects in product judgments when consumers are faced with irrelevant a novel and important domain - attribute weighing in consumer de- information. cision-making. Finally, our findings shed light on potential remedies In studies 1-3 we find that PA affects how decision-makers against bias in situations when people tend to underweight the impor- make desirability-feasibility trade-offs in consumer decisions. Usu- tance of feasibility attributes and overweigh irrelevant information. ally, decision-makers tend to overly contrite on the desirability considerations, often at the expense of feasibility considerations. Our

Advances in Consumer Research 999 Volume 45, ©2017 1000 / Exercise Your Mind – Physical Activity Alters Attribute Weighing in Consumer Choice REFERENCES Moore, Steven C., I-Min Lee, Elisabete Weiderpass, Peter T. Hopkins, Michael E., Caroline F. Davis, Michelle R. VanTieghem, Campbell, Joshua N. Sampson, Cari M. Kitahara, et al. (2016), Paul J. Whalen and David J. Bucci (2012), “Differential “Association of Leisure-Time Physical Activity With Risk of Effects of Acute and Regular Physical Exercise on Cognition 26 Types of Cancer in 1.44 Million Adults,” Jama Internal and Affect,”Neuroscience, 215, 59-68. Medicine, 176, 816-825. Kramer, Arthur F. and Kirk I. Erickson (2007), “Capitalizing on Morris, Jeremiah N., Heady, J. A., Raffle, P. A. B., Roberts, C. G. cortical plasticity: influence of physical activity on cognition and Parks, J. W. (1953), “Coronary Heart-Disease and Physical and brain function,” Trends in Cognitive Sciences, 11, 342- Activity of Work,” Lancet, 265, 1111-1120. 348. Nisbett, Richard E., Henry Zukier and Ronald E. Lemley (1981), Liberman, Nira and Yaacov Trope (1998), “The role of feasibility “The dilution effect: Nondiagnostic information weakens the and desirability considerations in near and distant future implications of diagnostic information,” Cognitive Psychology, decisions: A test of temporal construal theory,” Journal of 13, 248-277. Personality and Social Psychology, 75, 5-18. Raichlen, David A., Pradyumna K. Bharadwaj, Megan C. Fitzhugh, Liu, Wendy (2008), “Focusing on Desirability: The Effect of Karl A. Haws, Gabrielle-Ann Torre, Theodore P. Trouard et al. Decision Interruption and Suspension on Preferences,” (2016), “Differences in Resting State Functional Connectivity Journal of Consumer Research, 35, 640-652. between Young Adult Endurance Athletes and Healthy Meyvis, Tom and Chris Janiszewski (2002), “Consumers’ beliefs Controls,” Front Hum Neurosci, 10, 1-14. about product benefits: the effect of obviously irrelevant product information,” Journal of Consumer Research, 28, 618-635. Videography 2017

Luxurious Emirati Weddings: The Expenses, Pressures and Consequences Damien Arthur, Zayed University, UAE Sara Mohamed Al Marzooqi, Zayed University, UAE Nuha Nasser Salem Al Amri, Zayed University, UAE

Intended Contribution to Knowledge: This videography explores a ritual and site that due to religious and cultural norms is typically not for public broadcast. It is the first study to date to systematically analyze the expenses pressures and consequences associated with luxurious Emirati Weddings. It provides a unique insight into a closed culture where traditional rituals and values exist but have been warped by wealth and commercialization. Literature Foundations: This videography builds on the work of Russell Belk Rana Sobh and their colleagues (2011 2012 2013 2014) who explored how the concepts of hospitality privacy mimetic excess and gendered spaces are enacted by the people of the Arabian Gulf. Research Method: In-depth interviews and focus groups were undertaken with Emirati brides grooms mothers of grooms wedding attendees wedding planners and venue managers. As society’s norms encourage women to protect their image a major obstacle we overcame was depicting the female-only wedding reception without filming the occasion (as it is prohibited) and finding Emirati women willing to be interviewed on camera about the sensitive topic of marriage. To itemize the costs and quantify the pressures associated with hosting an Emirati wedding an online self-completion survey of 44 Emirati brides and 50 mothers of Emirati grooms was also conducted. Findings and implications: The omnipresent forces of culture religion and wealth have combined to normalize the commercialized and luxurious weddings of the indigenous people of the UAE. The average cost of hosting an Emirati wedding is now over US $185000. It is being driven largely by normative pressure to conform and avoid criticism that may impact family honor. Cultural values of hospitality and generosity also promote excess and are at odds with the Islamic value of modesty. Concerns about the cultural and societal impact exist however the population exhibit little will to change. Envisioning How New Technologies Can Fulfill Customers’ Unmet Needs Yaliang Chuang, Eindhoven University of Technology, The Netherlands Yu-Shan Athena Chen, Eindhoven University of Technology, The Netherlands Lin-Lin Chen, National Taiwan University of Science and Technology, Taiwan Yung-Hsun Chen, National Taiwan University of Science and Technology, Taiwan

Intended Contribution to Knowledge: Undeniably home is the most important environment for human beings. This study investigates customers’ unmet needs develops design concepts with new IoT technologies envisions the values for family members and reveals its impact on human life and interpersonal relationships. Literature Foundations: The videography contributes to the literature regarding human-centered design and the acceptance and diffusion of innovation. We demonstrate the contextual approach in finding customers’ unspoken needs from their living contexts. Research Method: Our focus is to reveal how future technology impacts interaction among household members. There are five folds of studies. First ten families (with 2-6 household members and living in 712-1300 m2 living space) were longitudinal followed and interviewed. Second employing the thematic analysis six critical needs in home environment were consolidated. Third through multidisciplinary workshops three value propositions were identified for generating a hundred ideas. Fourth the video prototyping was used to sketch the user experiences. Finally a focus group was conducted to gather target consumers’ feedbacks purchasing attitudes and marketing potentials. Findings and implications: Our first contribution is the five unmet needs for household markets: a house with emotional intelligence having an adaptive atmosphere singing songs everywhere as you wish collecting living memories and the assistance of parenting. These scenarios could assist markers and designers to scheme the future home environment. Then we consolidate two value propositions which enhance consumers’ well-being and life satisfaction. First the home environment should provide emotional fulfillments. Second it could bridge the parent-children communications and educate the young children to build their routines that are compatible to the parents. Those concepts were envisioned in the videography and validated by eight target customers participated in the focus group discussion. The smart home is expected to ease tensions and bring happiness to family members. Excessive Online Gaming: Inside the Routine of an E-Sports Fanatic Thiago Rafael Ferreira Marques, SENAC, Brazil Tania Veludo-de-Oliveira, Escola de Administração de Empresas de São Paulo da Fundação Getulio Vargas – FGV EAESP, Brazil

Although playing digital games online is an enjoyable experience literature reveals that playing online excessively in extreme cases may lead to symptoms commonly related to addiction such as alienation from the real world degradation of academic performance loss of sense of Time mood swings and so on (Chiu Lee & Huang 2004). Therefore there is a need to study the motivations that lead to excessive behavior with the aim to outline new strategies to prevent the abusive consumption of online games (Kuss & Griffiths 2012). The objective of this videography is to identify these motivations and analyze how they interact in the context of competitive online games (e-sports). A series of in-depth interviews (20) was conducted with players of Dota 2 and Counter strike who closely follow the competitive scene of these games. It was observed that much of the players’ experience is anchored in the social relations and friendship created within the game circle and by the competitive mindset of the players. The informants reported a high perception of flow state during the online gaming experience which corroborates the findings of previous studies (e.g.: Lee 2009). Flow’s perception was found to be a highly influential factor of the behavior in question. The informants also reported a strong identification with the professional players who serve as role models to them. The mimicry

Advances in Consumer Research 1001 Volume 45, ©2017 1002 / Videography of their role models’ gaming behavior as seen in the consumer doppelganger effect (Ruvio Gavish & Shoham 2013) further explains the excessive gaming. The videography contributes with rich insights to deal with abusive e-sports online gaming. It offers important subsidies for the creation of marketing cues which could drive consumers away from both the near-addictive or addictive stages of behavior (Martin et al. 2013) and closer to a healthy consumption habit. Searching for RNGesus: A Study on the Use of Randomization in Video Games Jacob Hiler, Ohio University, USA William Northington, Idaho State University, USA Laurel Cook, West Virginia University, USA

The discussion of randomization in marketing up until this point has focused primarily on its use as an experimental methodological tool. This study however focuses on studying intentional randomization in consumer experiences notably in the video game industry. The focus of this study is to explore both how randomization is employed by developers in video games as well as how it is experienced and perceived on the part of their consumers. As Kozinets (2015) suggests in Netnography Redefined more netnographic attention needs to be placed on video websites such as YouTube and Twitch rather than textual data in online forums and communities especially since many online communities and much of the discourse are moving to more video based discussions. Thus using this netnographic videography approach the filmmakers immersed themselves in over 25 hours of user-generated video content shared publicly on YouTube and Twitch and various user-generated text commenting on the nature of the random across various forums comment sections and reddit. Tomorrowland Festival: A Journey in Devirotopia Julie Masset, University of Namur, Belgium Alain Decrop, University of Namur, Belgium

Intended contribution: Each year the music festival ‘Tomorrowland’ gathers in Belgium about 180000 people who come from all over the world. A large number of festival-goers walk around with national flags accessories or fancy dresses which makes the festival unique. This research investigates consumers’ identity construction and expression through such material objects. Few studies have explored identity issues related to a music festival and the paraphernalia consumed during a festival. Literature: This videography relies on consumer research literature related to the meanings of objects (i.e. the role of material objects in consumers’ identity construction) and the consumption experiences (e.g. experiential marketing). Methodology: Through a naturalistic interpretive approach we interviewed 29 informants of both genders and of different ages nationalities educational levels and professional statuses. We also observed the festival and camping sites and the behaviors of festival-goers. Findings and implications: Emerging from data analysis and interpretation this videography documents various profiles of festival-goers from their material possessions (i.e. the patriot the devotee the child the bohemian girl and the provocateur) as well as the functions fulfilled by these possessions in their identity construction (i.e. national pride egological exhibitionism universal brotherhood and normalized deviation). We also propose two more original interpretations that enhance our contribution to the literature. First the consumption of paraphernalia used at Tomorrowland is typical of a cathartic release of inner tensions and emotions. The festival creates an “absorbing experience” that allows a “controlled decontrolling of emotions” in an increasingly disenchanted and ruled contemporary world. Second referring to Foucault’s concept of heterotopia the festival can be considered as a heterotopia of deviation because many festival-goers take benefit from Tomorrowland to regress and/or transgress societal rules and norms. The film illustrates a series of ritualized and commoditized deviant behaviors taking place in the fantasy and idle context of the festival. Becoming the Character: The Cosplayer Experience in a Con Stefânia Ordovás de Almeida, Pontíficia Universidade Católica do Rio Grande do Sul, Brazil João Pedro Fleck, Pontíficia Universidade Católica do Rio Grande do Sul, Brazil Rafael Mello, Pontíficia Universidade Católica do Rio Grande do Sul, Brazil

Intended Contribution to Knowledge: This study aims at contributing to the knowledge on the role played by the consumer when he produces elements of his consumption experience (WEIJO 2016) and to understand how this process might be used to contribute to their personal development (RYFF & SINGER 2008). Literature Foundations We know that many products can generate unique experiences that wouldn’t happen without their usage (e.g. surfboard GUEVARRA & HOWELL 2015) however with few exceptions (such as SEREGINA & WEIJO 2016) we know little about how the consumer can play an active role in taking this products to consumption environments and how this can create positive experiences that contribute to them as persons. - Research Method: The study involved a participant observation on 12 days of geek culture events and cons and also 14 interviews were videorecorded. Field notes and interviews were transcribed and read many times leading to the interpretation of results according to ideas of Spiggle (1994); McCracken (1988); and Arnould and Wallendorf (1994). –Findings and implications: As the cosplayers bring a product with them to the conventions they dedicate themselves to change their consumption experience generating a new set of behaviors and meanings. Many consume these events as a form of enjoyment however there are some the consume the events aiming at living a significant experience that can be also a personal challenge for them such as overcoming shyness or the lack of skills therefore the consumption ends up being a form of personal growth. Advances in Consumer Research (Volume 45) / 1003 Re-Assembling: Social Entrepreneurship, Motorcycles and Cast-Away Youth Joonas Rokka, EMLYON Business School, France Joel Hietanen, Aalto University, Finland Klaus Kangaspunta, Finland John Schouten, St. John University, Canada

“Re-assembling” is a videography that explores the notion of social entrepreneurship in the context of how ‘cast-away’ youth with little or no marketable professional skills can be brought back to working-life and re-connected with meaningful lives and sustainable ways of being. The film unfolds as a story about and around our social entrepreneur. Working at his “workshop” where bikes and motorbikes are re-assembled from abandoned lost or used parts often discovered from trash or parts re-circulated in second-hand bike market. The parts are carefully re-worked re-connected and put together as unique beautiful manually crafted bikes that are then sold or traded forward. But he not only re-assembles and sells bikes. He has a broader mission and calling that links his workshop with a social cause. For three years he has started to bring along youngsters often from difficult backgrounds to his workshop to learn what he calls ‘basic life skills’. It is this ongoing re-assembling – of both material objects but also humans – that the film examines and its implications for conceiving potential new forms of social entrepreneurship. We thus seek to illuminate how the new form entrepreneurship performed in one field – such as the practice of motorcycle construction – via accumulation of resources (material social networks and skills) may engender potential for the possibility for translating those resources to simultaneously addressing a social cause – here by supporting the cast-away youngsters. In this way we hope to increase understandings about the work and challenges of such social entrepreneurship. Stealing from the Rich Stefan H. Szugalski, Stockholm School of Economics, Sweden Jonas Colliander, Stockholm School of Economics, Sweden Magnus Söderlund, Stockholm School of Economics, Sweden Sofie Sagfossen, Stockholm School of Economics, Sweden

Intended Contribution to Knowledge: We explore the effects bad behavior might have and look at this from a marketplace morality perspective. Researchers have mainly focused on good and bad behavior as two different acts. With this videography we expect to widen the knowledge on morality in the marketplace and specifically add perspectives on self-interest. We extend knowledge by revealing how economical and other self-interests contradict each other. We discuss personal desires and the effects the experience of winning may have on a group of people to the extent that it might impact a whole town. Literature Foundations: The theoretical foundation of the study is the assumption that a marketplace of morality is a market of morality inputs and outputs (Dunfee 1988). Here we study a case in which a person donated money (good behavior) to a sports organization which were obtained illegally (bad behavior) and investigate the possible positive effects of it. What really makes this case interesting from a research perspective is that the person that carried out the dishonest behavior was also the one who lost money as the funds were embezzled from his family business. Research Method: The empirical base is an interview with the person that carried out the illegal act (the input). We also conducted interviews with people living in the town where it all took place. Together with secondary data (articles from 2003-2010) this gave us a ground to explore the impact a “winning culture” might have on collective well-being (the output). Findings and implications: We extend the knowledge on the possible positive effects immoral behavior may have on the marketplace and about value-transfer in the marketplace. This story sheds light on the impact of winning and demonstrates how experiences can increase well-being in a way that products do not. 1004 / Videography Roundtable Summaries

Roundtable The Evolution of Signals

Participants: Russ Belk, York University, Canada Silvia Bellezza, Columbia University, USA Jonah Berger, University of Pennsylvania, USA James R. Bettman, Duke University, USA David Gal, University of Illinois at Chicago, USA Anat Keinan, Harvard University, USA Amna Kirmani, University of Maryland, USA Ayalla Ruvio, Michigan State University, USA Yajin Wang, University of Maryland, USA Morgan K. Ward, Emory University, USA Caleb Warren, University of Arizona, USA

Consumers leverage the symbolic nature of products to signal unobservable characteristics such as identity or status. Due to their symbolic nature the meaning and hence use of signals is constantly evolving. This roundtable discussion explores open questions regarding the antecedents and consequences of signal evolution.

Roundtable Expanding the Horizons of Videographic Consumer Research

Participants: Russell Belk, York University, Canada Robert Kozinets, University of Southern California, USA Brownlie Douglas, University of Dundee, UK Paul Henry, University of Sydney, Australia Joel Hietanen, Aalto University, Finland Eric Li, University of British Columbia, USA Jacob Ostberg, Stockholm Business School, Sweden John Schouten, University of St. John, Canada Baptiste Cléret, University of Rouen, France Anastasia Seregina, Aalto University, Finland

The purpose of this roundtable is to map out and explore the future directions for videography in consumer research. While only limited ontological epistemological and methodological accounts readily exist to assist videographers this session intends to explore potential new pathways for video-based theorizing and the production of consumer research knowledge.

Participants: Jonathan Z. Berman, London Business School, UK James R. Bettman, Duke University, USA Cynthia E. Cryder, Washington University, USA Elizabeth W. Dunn, University of British Columbia, Canada Gavan J. Fitzsimons, Duke University, USA Cassie Mogilner Holmes, University of California Los Angeles, USA Michael I. Norton, Harvard University, USA Jenny G. Olson, Indiana University, USA Colbey Emerson Reid, North Carolina State University, USA Scott Rick, University of Michigan, USA Avni M. Shah, University of Toronto, Canada Stephen A. Spiller, University of California Los Angeles, USA Abigail B. Sussman, University of Chicago, USA Kathleen D. Vohs, University of Minnesota, USA Lisa A. Cavanaugh, University of British Columbia, Canada

As in everyday life discussions of consequential financial decisions and social processes have been largely separate in consumer behavior and psychology. This roundtable will consider how work on decision-making financial behavior social dynamics and interpersonal relationships can inform one another to better understand how these domains interact in consumer’s lives.

Roundtable Everyday Consumer Aesthetics

Participants: Lauren Block, Baruch College, USA Laura Peracchio, University of Wisconsin - Milwaukee, USA Claudia Townsend, University of Miami, USA Mario Pandelaere, Virginia Tech, USA Ravi Mehta, University of Illinois at Urbana-Champaign, USA JoAndrea (Joey) Hoegg, University of British Columbia, Canada Maureen Morrin, Temple University, USA Aradhna Krishna, University of Michigan, USA Maura Scott, Florida State University, USA Luca Cian, University of Virginia, USA Henrik Hagvtedt, Boston College, USA Adriana Madzharov, Stevens Institute of Technology, USA Kelly Herd, Indiana University, USA Maura Scott, Florida State University, USA Martin Reimann, University of Arizona, USA Theodore Noseworthy, York University, USA Xiaojing Yang, University of Wisconsin – Milwakee, USA Xiaoyan Deng, Ohio State University, USA Yuwei Jiang, Hong Kong Polytechnic University, Hong Kong Lei Jia, Ohio State University, USA Rita To, University of Houston, USA Zhe Zhang, University of Houston, USA Tim Doering, University of Michigan, USA Julio Sevilla, University of Georgia, USA

Everyday consumer aesthetics is defined as “non-art non-nature aesthetic experiences that are diverse and dynamic and result in specific consumer actions (e.g. purchasing) and consumption behaviors” (Patrick 2016). The roundtable will discuss the current state of the literature related research areas and future directions for this sub-field of consumer behavior. Advances in Consumer Research (Volume 45) / 1007 Roundtable Theoretical Advances in Consumer Neuroscience: How Affective, Cognitive, and Social Neuroscience Informs Consumer Behavior (and Vice Versa)

Participants: Clark Cao, University of Arizona, USA Adam Craig, University of Kentucky, USA Raquel Castaño, Tecnológico de Monterrey, Mexico Moran Cerf, Northwestern University, USA Angelika Dimoka, Temple University, USA Bill Hedgcock, University of Iowa, USA Hilke Plassmann, INSEAD, France Martin Reimann, University of Arizona, USA Ale Smidts, Erasmus University Rotterdam, The Netherlands Carolyn Yoon, University of Michigan, USA

During the roundtable participants will debate theoretical advances in consumer neuroscience especially how concepts from affective cognitive and social neuroscience can inform consumer research and vice versa. The roundtable aims to develop directions for theory development and testing in the field of the neurophysiology of consumption.

Roundtable Conspicuous Consumption Revisited in the Digital Era

Participants: David Dubois, INSEAD, France Paurav Shukla, Essex University, UK Valentina Clergue, HEC Lausanne, Switzerland Bruno Kocher, HEC Lausanne, Switzerland Leif Brandes, Warwick University, UK Martin Liu, Nottingham University, Ningbo Campus, China Teck Y. Eng, University of Southampton, UK Julia Wolny, University of Southampton, UK Chunyan Xie, Western Norway University of Applied Sciences, Norway Lily Wang, Zhejiang University, China Elodie Caucigh, University of Vienna, Austria Jeff Lee, NYU Shanghai, China Nailya Ordabayeva, Boston College, USA

Digital technology including social media challenge fundamental premises of luxury brands including exclusivity status authenticity and sensory experiences. Digital luxury experiences differ from their offline counterparts and brought forward terms like “webmosphere” or “luxemosphere” indicating new approaches. Little is known about the underlying mechanisms manifested in these new phenomena.

Roundtable What is a Consumption Experience?

Participants: Richard Lutz, University of Florida, USA Hope Jensen Schau, University of Arizona, USA David Mick, University of Virginia, USA Deborah Macinnis, University of Southern California, USA Cassie Mogilner Holmes, University of California Los Angeles, USA Rebecca Hamilton, Georgetown University, USA Markus Giesler, York University, Canada Robin Coulter, University of Connecticut, USA Eileen Fischer, York University, Canada Linda Price, University of Oregon, USA

Consumption experience is the heart of consumer behavior and value creation. This roundtable will explore several questions related to consumption experiences. What is a consumption experience and its most important dimensions? How do we study each dimension? How can we develop a more comprehensive view of consumption experience? This session integrates diverse perspectives and methodologies to guide research investigating how consumption experience drives value creation. 1008 / Roundtable Summaries Roundtable Charting the Future of the Transformative Consumer Research Movement

Participants: David Mick, University of Virginia, USA Cornelia Pechmann, University of California Irvine, USA Maura Scott, Florida State University, USA Martin Mende, Florida State University, USA Janet McColl-Kennedy, University of Queensland, Australia Lucie Ozanne, University of Canterbury, New Zealand Chris Blocker, Colorado State University, USA Andres Barrios, Universidad de los Andes, Colombia Laurel Anderson, Arizona State University, USA Jeff Murray, University of Arkansas, USA Lauren Block, Baruch College, USA Linda Price, University of Oregon, USA Madhu Viswanathan, University of Illinois, Urbana-Champaign, USA Leonard Lee, National University of Singapore, Singapore Beth Vallen, Villanova University, USA Karen Winterich, Penn State, USA Laura Peracchio, University of Wisconsin Milwaukee, USA Rodrigo Castilhos, Universidade Federal do Rio Grande do Sul, Portugal Samuelson Appau, University of Melbourne, Australia L. Lin Ong, Cal Poly Pomona, USA Marcus Phipps, University of Melbourne, Australia Shikha Upadhyaya, California State University Los Angeles, USA Emily C. Tanner, West Virginia University, USA Elizabeth Crosby, University of Wisconsin, USA Christopher Berry, Colorado State University, USA Sterling Bone, Utah State University, USA Daniele Mathras, Northeastern University, USA Kelly Haws, Vanderbilt University, USA Zafeirenia Brokalaki, Kings’ College London, UK

Transformative Consumer Research leaders from the advisory committee or past conference track chairs organize five task force discussions around key challenges and opportunities facing the movement. Each task force will present initial recommendations and then seek questions and input from ACR roundtable attendees for a strategic planning report. Advances in Consumer Research (Volume 45) / 1009 Roundtable Toward an Integrated Understanding of Resource Scarcity

Participants: Andrea Morales, Arizona State University, USA Angela Lee, Northwestern University, USA Anthony Salerno, University of Cincinnati, USA Anuj Shah, University of Chicago, USA Brent McFerran, Simon Fraser University, Canada Chiraag Mittal, Texas A&M University, USA Danit Ein-Gar, Tel Aviv University, Israel Darren Dhal, University of British Columbia, Canada Debora Thompson, Georgetown University, USA Deborah Roedder John, University of Minnesota, USA Eesha Sharma, Dartmouth College, USA Juliano Laran, University of Miami, USA Julio Sevilla, University of Georgia, USA Kirk Kristofferson, Arizona State University, USA Lan Nguyen Chaplin, University of Illinois at Chicago, USA Liat Levontin, Israel institute of Technology, Israel Meryl Gardner, University of Delaware, USA Rebecca Hamilton, Georgetown University, USA Rebecca Ratner, University of Maryland, USA Sarah Hill, Texas Christian University, USA Kristina Durante, Rutgers University, USA

This roundtable aims to encourage constructive discussion and create a space for researchers interested in resource scarcity to start integrating the different definitions operationalizations and findings from extant scarcity research into a coherent framework in order to identify gaps requiring further attention and share ideas that will stimulate future collaborations. 1010 / Roundtable Summaries Working Papers

17-D: Friends with Benefits: Social Support and Pain of Payment

Deborah Abrams, Georgia Tech, USA Samuel Bond, Georgia Tech, USA Michael Lowe, Georgia Tech, USA

This research demonstrates how social presence through its effect on social support can attenuate pain of payment. Studies 1 and 2 show a direct negative effect of social presence on pain of payment as well as a mediating effect of social support.

13-J: The Effect of Stress on Consumers’ Private Information Disclosure

Sinem Acar-Burkay, University College of Southeast Norway, Norway Bob M. Fennis, University of Groningen, The Netherlands

Consumer privacy in the age of big data is an important research topic for consumer researchers. In two studies we examine how stress affects private information disclosure. Results show that stress increases consumers’ likelihood of answering highly-sensitive and even in- criminating questions affirmatively while potentially making them susceptible to privacy risks.

20-U: Who are the Influentials? Social Media, Opinion Leadership and New Product Adoption

Duygu Akdevelioglu, University of California Irvine, USA Selcan Kara, University of Massachusetts, Dartmouth, USA

This research examines how innovativeness and extraversion influence self-reported and sociometric opinion leadership in social media which affect new product adoption. Findings contribute to extant work on new product adoption by showing that self-reported leadership and sociometric leadership are distinct characteristics because they act differently in their nomological networks.

8-E: Jeitinho Brasileiro: Understanding Financial Vulnerability from a Cultural Perspective

Rafaela Almeida Cordeiro, Escola Superior de Propaganda e Marketing, Brazil Nancy Wong, University of Wisconsin - Madison, USA Mateus Ponchio, Escola Superior de Propaganda e Marketing, Brazil

The current research provides a unique context in showing how financial vulnerability is jointly shaped by weak policies and consumer values. It contributes to market system dynamics in explaining how culture and social capital shape consumer resistance and coping strategies in the Brazilian marketplace.

9-FF: Emotional Labor and Emotional Intelligence in Healthcare Delivery Organizations: The Case of Medicaid Recipients

Eklou Amendah, Southern New Hampshire University, USA Adrienne Hall-Phillips, Worcester Polytechnic Institute, USA

For healthcare providers serving Medicaid recipients emotions play an important role when managing beneficiaries and recipients. The study aims to demonstrate that service recipient characteristics and work environment challenges moderate the relationship between emotional intelligence and emotional labor when Medicaid beneficiaries are being served by healthcare organization employees.

Jiyoon An, University of Rhode Island, USA Daniel Sheinin, University of Rhode Island, USA

Drugs and dietary supplements are now often available in the same distribution channels. Drugs are heavily-regulated medical products while dietary supplements are lightly-regulated food products. This paper explores whether anthropomorphized brand role (servant versus partner) changes perceptions about these products and intentions to engage in healthy behaviors.

1-F: Grotesque Imagery Enhancing Persuasiveness of Luxury Brand Advertising

Donghwy An, Hongik University, South Korea Chulsung Lee, Hongik University, South Korea Janghyun Kim, Korea Christian University, South Korea Nara Youn, Hongik University, South Korea

The current research examines consumers’ perceived fit between grotesque imagery and luxury branding. We showed that the fit between grotesque imagery and luxury brand advertising facilitates transportation which in turn enriches brand experience and increases purchase intention of the featured product.

1-A: Art Appreciation Opens Affirmed Minds to Cultural Diversity

Donghwy An, Hongik University, South Korea Boram Lee, Hongik University, South Korea Nara Youn, Hongik University, South Korea

Appreciating artworks enhanced endorsement of cultural diversity for individuals with high self-affirmation (Study 1) and for those with authentic pride but not for those with hubristic pride (Study 2). The effect of art appreciation was significant for narcissistic individuals only when their openness to cultural diversity was publicized (Study 2).

13-G: New with Tags: Consumer and Brand Relationships in Consumer-to-Consumer Buy/Sell/Trade Groups on Social Media

Catherine Armstrong Soule, Western Washington University, USA Sara Hanson, University of Richmond, USA

This paper introduces and explores the emerging and impactful phenomenon of branded consumer-to-consumer buy/sell/trade groups on social media. These groups are independently organized marketplaces where consumers buy and sell one focal brand exclusively. Through in-depth interviews field data scraped from Facebook and survey data we identify brand and retailing implications.

14-I: Signaling Nothing: Conspicuous Anti-Consumption in Demarketing Contexts

Catherine Armstrong Soule, Western Washington University, USA Tejvir Sekhon, Western Washington University, USA

This paper explores the link between status-signaling and demarketing. The anti-consumption resulting from demarketing strategies does not have a public display of virtuous behavior. We test whether an observable signal can motivate anti-consumption by providing status-signaling opportunity and whether this signal benefits the anti-consumer as well as the demarketing brand. Advances in Consumer Research (Volume 45) / 1013 19-A: The Interactive Effect of Type and Framing of Discounts on Consumers’ Preference

Amin Attari, University of Kansas, USA Promothesh Chatterjee, University of Kansas, USA Surendra Singh, University of Kansas, USA

This article investigates consumers’ preference for probabilistic versus sure discounts across two frames the amount of discount frame and the reduced price frame. We show that consumers’ preference between the sure and probabilistic discounts can reverse depending on the type of frame used to present the discount.

13-K: The Fit Between Depth of Self-Disclosure and Self-Construal on Consumer Evaluations in Peer-to-Peer Markets

Sumitra Auschaitrakul, University of the Thai Chamber of Commerce, Thailand Ashesh Mukherjee, McGill University, Canada Dan King, University of Texas Rio Grande Valley, USA

Peer-to-Peer markets (P-to-P) are platforms where individual sellers exchange products with individual buyers. Based on a fit-fluency framework the present research shows that depth of disclosure in peer-to-peer markets has a positive effect on product evaluation when self-construal is interdependent but this positive effect is eliminated when self-construal is independent.

9-GG: Narrative Engagement: A Model for Suicide Prevention

Caroline Graham Austin, Montana State University, USA Sarah Keller, Montana State University, USA

This paper explores one community-based approach to addressing suicide. Narrative engagement encourages people to collectively share personal stories yielding information about shared perceptions attitudes and beliefs. Such community engagement shines light into “dark closets” reducing stigma and isolation. Identifying communal strengths and weaknesses builds resiliency and identifies potential resources.

11-J: Moral Legitimation of Hybrid Consumption Practices: A Study of Gluten-Free Lifestyles

Ana Babic Rosario, University of Denver, USA

This study explores how consumers legitimize hybrid consumption practices such as gluten-free eating to support their identity work. Based on netnographic interview and archival data several consumer strategies for moral legitimation were identified. The insights extend the work on consumer and practice legitimation to the context of hybridized cultural forms.

14-U: Seeing Through Consumer: Predicting Consumer Preferences and Choices by using Deep Learning Neural Networks

Marija Banovic, Aarhus University, Denmark Vladan Arsenijevic, Seven Bridges Genomics, Serbia

Training neural networks with consumer data can critically change marketing practice. This research shows that consumer choices can be predicted by usage of deep learning neural networks. Apart from prediction of what consumer is most likely to select we can optimize the product to fit consumer’s preferences. 1014 / Working Papers 5-M: Saved by the Past? Activated Disease Threat Promotes Consumer Preferences for Nostalgic Products

Dovilė Barauskaitė, ISM University of Management and Economics, Lithuania Justina Gineikienė, ISM University of Management and Economics, Lithuania Bob M. Fennis, University of Groningen, The Netherlands

Consumer research experienced a surge in interest in when and why consumers prefer nostalgic products. The present work complements this literature by examining the role of a yet untested antecedent for nostalgic consumption – salient disease threat. Across two studies we show that activated disease threat promotes heightened consumer nostalgic preferences.

7-G: How the Uniqueness of Goods Influences Consumers’ Response to Price Increase for Experiences versus Objects

Wilson Bastos, Catholic University of Portugal, Portugal

This research shows that consumers react more favorably to a price increase when it is associated with an experiential good than a material one. Further it examines five potential explanations (closeness to the self conversational value impression management social relatedness and purchase uniqueness) and reveals uniqueness as the mechanism.

7-C: Conceptualizing the Digital Experience in Luxury

Wided Batat, University of Lyon 2, France Olivier Sinai, Birkbeck, University of London, UK

This research introduces a theory of digital luxury experience conceptualizing what it is and identifying its drivers. This is of importance to luxury consumption as it offers a framework to understand digital luxury an empirical phenomenon which has gained prominence in the last 10 years and has remained poorly understood.

7-H: Identity Construction Through Destination Experience: An Introspective Approach

Wided Batat, University of Lyon 2, France

This research draws on a subjective personal introspection (SPI) approach and Breakwell’s identity process theory (IPT) principles to show how elements from different cultures are performed by an individual to form a unique patchwork identity and how this patchwork identity will contribute to deepen consumer experience of a cultural destination.

11-I: Effects of the Qualification of Potentially Misleading Claims on Production Characteristics on Buying Intention

Tino Bech-Larsen, Aarhus University, Denmark George Tsalis, Aarhus University, Denmark Trine Moerk, Aarhus University, Denmark Klaus G. Grunert, Aarhus University, Denmark

This consumer (n=1420) study analyzes the qualification of potentially misleading statements about production processes based on conjoint ratings of taste health and animal welfare qualities and buying intentions. Deception potential is revealed by mediation tests of taste health and AW ratings on the effects of conjoint attributes on buying intentions. Advances in Consumer Research (Volume 45) / 1015 14-O: The Neuroscience Viewpoint on the Interaction Effect between Persuasive Arguments and Celebrity on Attitude Change

Lien-Ti Bei, National Chengchi University, Taiwan Li-Keng Cheng, National Chengchi University, Taiwan Wei-Chin Hsu, National Taiwan University of Science and Technology, Taiwan Yu-Shan Athena Chen, Technische Universiteit Eindhoven, The Netherlands

This study investigated the interaction effect between arguments and celebrity on attitude. The results of traditional and fMRI experiments showed that heuristic arguments with celebrity required more cognitive resources than strong and weak arguments. Also consumers generated stronger emotion reaction toward celebrities with heuristic arguments than strong and weak arguments.

9-CC: Will Consumers Accept Foods Made from Surplus Ingredients?

Siddharth Bhatt, Drexel University, USA Jeonggyu Lee, Drexel University, USA Jonathan Deutsch, Drexel University, USA Hasan Ayaz, Drexel University, USA Benjamin Fulton, Drexel University, USA Rajneesh Suri, Drexel University, USA

To address the global food waste problem researchers have proposed creating foods from ingredients that would otherwise be wasted. This research examines consumers’ perceptions of - Value Added Surplus Products which are foods made from surplus ingredients to conclude that consumers indeed accept these as a new category.

12-F: The Motivation to Exclude Others and its Impact on Brand Commitment

Samantha Bittner, Florida State University, USA Tatiana Fajardo, Florida State University, USA

This research examines exclusion in the context of brand communities. The authors demonstrate that in communities of high (low) personal relevance consumers are more (less) motivated to exclude others. Furthermore in communities of high (low) personal relevance consumers report higher brand commitment after excluding (including) another from the brand community.

4-Q: The Use of Ethnically Ambiguous Models in Advertising to Preserve Consumer Well-being in an Organic Multicultural Market

Wim Booyse, University of Pretoria, South Africa Samantha Swanepoel, University of Pretoria, South Africa

Our research compares a multiculturated consumers’ feeling of well-being when evaluating advertisements with differing variations of ethnic targeting. Consumers within a multicultural market (South Africa) have a greater feeling of consumer wellbeing when evaluating an advertisement containing ethnically ambiguous models compared to models of multiple ethnicities or individual ethnic models.

9-T: Savoring Stress: Can Feeling Stressed Reduce the Rate of Satiation?

Benjamin Borenstein, University of Miami, USA Juliano Laran, University of Miami, USA Luke Nowlan, University of Miami, USA

In this research we propose that stress reduces the rate of satiation to hedonic consumption experiences. In order to restore control stressed individuals engage more with the activities they are currently pursuing. As a result they take longer to satiate to music (study 1) and food (study 2). 1016 / Working Papers 20-J: Promotional Reviews, Social Influence, and the Information Content of Online Reviews

Leif Brandes, University of Warwick, UK Rosa Cacabelos, University of Zurich, Switzerland Egon Franck, University of Zurich, Switzerland

Does social influence during the review provision process reduce the information content of reviews when individuals are exposed to promotional reviews? We show that social influence reduces the evaluation gap between products of different quality levels but that this reduction is insufficient to make average review scores uninformative.

1-G: I Want the Best, So Give Me All the Bells and Whistles: How Maximizers versus Satisficers Evaluate Feature-Rich Products

Daniel Brannon, University of Northern Colorado, USA Brandon Soltwisch, University of Northern Colorado, USA

We find that maximizers (vs. satisficers) give more favorable evaluations to feature-rich products indicating that they are less likely to anticipate feature fatigue. Underlying this relationship is a dual process whereby maximizers perceive feature-rich products as status signals as well as overestimate how much they will use the additional features.

11-S: When Do Ethical Consumers Elicit Inspiration Instead of Denigration?

Claire Brouwer, Universitat Pompeu Fabra, Spain Jan Willem Bolderdijk, University of Groningen, The Netherlands Gert Cornelissen, Universitat Pompeu Fabra, Spain

This work identifies a boundary condition under which ethically motivated consumers elicit positive responses from observers rather than typically negative ones. We show that exemplars have the potential to harvest applause and can inspire others to emulate but only if there is sufficient psychological distance between themselves and potential observers.

20-K: Provision of Sales Volume Information Influences Consumers’ Preferences Via Perceived Purchase and Word-of-Mouth Popularity

Adrian Camilleri, RMIT University, Australia

Across three experiments we show that when consumers are presented with both the number of online reviews and the number of sales many of them use the ratio between these numbers – the reviewer percentage – to infer a product’s “word-of-mouth popularity” which is distinct from a product’s “purchase popularity”.

2-A: Shante you Stay/ Sashay Away: Approaching Language Within Brand Assemblages

Mario Campana, Goldsmiths, University of London, UK Katherine Duffy, University of Glasgow, UK Alessandro Caliandro, Middlesex University, UK

This research extends brand assemblage theory to contribute to understanding how language assembles and evolves in brand niche communities. We question how consumers of RuPaul’s DragRace create and circulate a cultural product then we explore how this is scaled-up and mainstreamed moving from a brand minor to a brand major. Advances in Consumer Research (Volume 45) / 1017 9-L: Healthy or Happy? How Awe Enhances Consumers’ Preferences for Healthy Products

Fei Cao, Renmin University of China, P.R. China Xia Wang, Renmin University of China, P.R. China Ze Wang, University of Central Florida, USA

Three studies demonstrate that experiencing awe (vs. neutral emotion) enhances consumers’ preferences for healthy products (vs. unhealthy products). The effect is driven by analytic processing and is stronger for consumers with chronic affective (vs. cognitive) information processing style.

7-K: Negotiated Outsourcing – Towards an Understanding of Market Tension in Enduring Consumption Experiences

Flavia Cardoso, Universidad Adolfo Ibáñez, Chile Pilar Rojas-Gaviria, Pontificia Universidad Católica de Chile, Chile Daiane Scaraboto, Pontificia Universidad Católica de Chile, Chile

We contribute to a recent stream of research on consumer outsourcing by examining enduring (i.e. long-term transformational) consumption experiences. We identify cultural and social sources of tension related to outsourcing in the schooling market and map the emergence of a parallel market in which these issues are negotiated and resolved.

2-R: Oxytocin Increases Trust for Familiar and Familial Brands

Lilian Carvalho, Fundacao Getulio Vargas, Brazil Gad Saad, Concordia University, Canada Eliane Brito, Fundacao Getulio Vargas, Brazil

We designed two experiments to investigate the relationship between oxytocin and brand trust. The first study found that both known and unknown brands can reap the effects of oxytocin. In the second study we investigated if oxytocin had effects for both social vs. status stimuli and also found significant effects.

17-A: Consumer Perceptions of Social Robots

Noah Castelo, Columbia University, USA Bernd Schmitt, Columbia University, USA

Robots are being used in a range of consumption settings. We explore how consumers’ perceptions of robots depend on the robots’ memory and emotionality. Using a social robot called Nadine we find that emotion and memory interact in shaping consumers’ perceptions of the robot.

5-Q: The Effect of Working with Ingroup versus Outgroup Members on Affective Forecasting Bias

Lorenzo Cecutti, The Chinese University of Hong Kong, China Jessica Kwong, The Chinese University of Hong Kong, China

Normally individual’s emotionality is considered detrimental to shared goals. However the present findings show that in group settings affective forecast bias can be functional to goal pursuit. We found that when working with ingroup members participants’ affective forecast motivated goal persistence. Yet this pattern reduced when working with outgroup members. 1018 / Working Papers 3-F: Pride Shapes Decisions through Construal Level: The Effect of Authentic vs. Hubristic Pride on Construal Level

Melis Ceylan, Koc University, Turkey Nilüfer Aydınoğlu, Koc University, Turkey Selin Atalay, Frankfurt School of Finance and Management, Germany

Pride is a pervasive emotion. Thus it is critical to understand how this emotion affects how consumers construe information and make subsequent choices. This research demonstrates that individuals experiencing authentic (hubristic) pride adopt lower- (higher-) levels of construal. This result contributes to research in emotions and judgment & decision making.

10-C: Calendar Framing Affects Perception and Anticipation of Sunday

Rebecca Chae, University of Michigan, USA Christine Kang, California State University Long Beach, USA

We examine the effects of calendar framing on time perception and anticipatory emotions. We demonstrate how the location of Sunday on a calendar (i.e. at the beginning vs. the end of the week) influences perception and anticipation of Sunday. We further discuss its implications on marketing communications.

5-O: The Effect of Antecedents of Sadness on Brand Evaluation

Luxi Chai, University of Kansas, USA Yexin Jessica Li, University of Kansas, USA

The current paper contributes to research on emotion in consumer attitudes and behaviors by showing that when consumers experience sadness from status loss they value brand competence and rate competent brands more desirable than warm brands. Ratings of brand traits do not differ when consumers experience sadness from social loss.

9-A: “I Reward Myself with Salad!” Framing Healthy Food Option as a Reward Increases Purchasing Intention

Luxi Chai, University of Kansas, USA Marija Grishin, University of Kansas, USA

The current research examines the impact of gratification-enhancing framing on healthy food-related decision making by demonstrating that framing the healthy (vs. unhealthy) food option as a reward increases the rewarding feeling and purchase intention.

16-A: Adding a Charitable Donation to a Product or Adding a Product to a Charitable Donation? How Much You Can Raise Depends on How You Frame the Case

Chia-Chi Chang, National Chiao-Tung University, Taiwan Po-Yu Chen, National Chiao-Tung University, Taiwan

For charity sales hedonic products will solicit more help than utilitarian products. For donation with a thank-you gift adding a charitable donation to utilitarian products tend to be more effective in soliciting more help than adding the same donation to hedonic products.

19-D: Studying Bets and Hedges in Simple Lotteries

Subimal Chatterjee, SUNY Binghamton, USA Satadruta Mookherjee, SUNY Binghamton, USA

We use a simple lottery to test if consumers value a hedge against losing in any way that is different from how they value the risk and test some potential explanations that might drive these valuations. Advances in Consumer Research (Volume 45) / 1019 3-H: Through the Looking Glass: Role of Construal Level on Description-Intensive Reviews

Swagato Chatterjee, Institute for Financial Management and Research, India Aruna Divya T, IIM Ahmedabad, India

Focus on consumer engagement has led service providers to explore contextual factors influencing consumers’ satisfaction. In this paper we draw insights from Construal Level Theory to identify the conditions when own vs. others’ experiences along with Process vs. Outcome attributes of services become more important in overall service evaluation.

14-M: The Making of a Successful Advocate: How Incentivizing Consumer Referrals Affects Persuasion

Lauren Cheatham, Stanford University, USA Zakary Tormala, Stanford University, USA

When it comes to consumer product and service referrals do incentives work? We find that there is an important tradeoff such that incentivizing referrals increases reach but undermines the effectiveness of the advocacy. We provide evidence for this tradeoff being driven by a shift in intrinsic motivation.

4-G: Examining Tourists’ Intentions to Revisit Religion Processions: The Role of Religiosity and Flow Experience

Annie Chen, University of West London, UK Norman Peng, University of Westminster, UK

The purpose of this research is to examine the factors that contribute to tourists’ intentions to re-participate in religious processions. The results from 109 participants show flow experience positively affects revisit intentions. Addition religiosity positively moderates the relationship between flow experiences and revisit intentions.

5-B: Coping with On-going Terror and Threats: Resilience, Place Attachment, and Behaviors

Ning (Chris) Chen, University of Canterbury, New Zealand Rohail Ashraf, Lahore University of Management Sciences, Pakistan

This study looks at regions in Pakistan where people live in constant terror of terrorism political unrest and threats of death and aim to understand how people cope with the extreme situations how their attachment to the place evolves and changes and how people’s perceptions and ideologies on consumption shift.

6-A: Chooser or Suitor: The Effects of Mating Cues on Men’s vs . Women’s Reaction to Brand Rejection

Rui Chen, Xiamen University, China Chun-Ming Yang, Ming Chuan University, Taiwan Hao Shen, Chinese University of Hong Kong, China

Once mating motives are activated males will respond to rejection less negatively than they otherwise would whereas this effect does not occur for females. Two studies examine the marketing implications of these effects. 1020 / Working Papers 7-M: When a Minor Problem Causes a Major Trouble for Experiential Purchases

Qihui Chen, Peking University, China Jingjing Ma, Peking University, China

Although past research has consistently shown that experiential purchases produce higher satisfaction than material purchases this research shows that experiential purchases could lead to lower satisfaction when these purchases are mixed-valence purchases. This is because consumers evaluate experiential purchases based on feelings; a minor problem could cause a major dissatisfaction.

5-P: The Effect of Nostalgia Advertising on Brand Loyalty

Jia Chen, University of Illinois at Chicago, USA Jingjing Ma, Peking University, China

Prior research showed that nostalgic ads of certain brands improve consumers’ attitude towards these brands. Here we propose that if a brand has made inconsistent changes of its positioning in the past nostalgic ads can also result in decreased consumer loyalty. We speculate that self-brand connection might be a mechanism.

8-F: Or Else! The Positive Effect of an Advisor’s Ultimatum on Perceived Expertise

Shirley Shuo Chen, University of Alberta, Canada Kyle B Murray, University of Alberta, Canada

This research examines how advice given with an ultimatum affects perceived expertise in the context of financial decision-making. In a series of three studies we find that perceptions of expertise increase when an advisor threatens to provide no future advice if the current recommendation is not accepted.

10-E: An Institutional Perspective of Country-of-Origin and Its Effects on Consumer Purchase Preference

Shijiao Chen, University of Otago, New Zealand Damien Mather, University of Otago, New Zealand Hongzhi Gao, Victoria University of Wellington, New Zealand John Knight, University of Otago, New Zealand

This paper conceptualizes country-of-origin (COO) cues as indicators of country institutions and tests the effect of COO cues on consumers’ perceptions and purchase preference in the Chinese food market. This paper includes a survey experiment and a choice modelling experiment and offer implications for international marketing researchers and practitioners.

20-T: When Too Many Friends Inhibit Talking: The Impact of Social Roles on Social Sharing Online

Yu-Jen Chen, Lingnan, Hong Kong

We study how social roles affects online sharing behavior. We propose that as the number of friends increases consumers may perceive role conflict due to the perceived difficulty in managing diverse role expectation. As distinct social roles are made salient they may inhibit online sharing behavior.

1-B: Cross-Cultural Differences in Aesthetic Judgments of Products: The Role of Cognitive Styles

Tseng-Ping Chiu, University of Michigan, USA Carolyn Yoon, University of Michigan, USA Shinobu Kitayama, University of Michigan, USA

This study investigates cross-cultural differences in consumers’ aesthetic judgments of products in a matching versus non-matching context. Our findings suggest that those with holistic cognitive styles (East Asians) are more tolerant of object-context mismatch than those with more analytic styles (European Americans). Advances in Consumer Research (Volume 45) / 1021 18-A: Green by Itself, But Brown in Comparison: The Influence of Sustainability Disclosures at the Shelf

Yoon-Na Cho, Villanova University, USA Robin Soster, University of Arkansas, USA Scot Burton, University of Arkansas, USA

Consumers receive little or no consistent brand level sustainability information. Drawing from comparative processing and information disclosure literatures we conduct (1) a retail laboratory experiment and (2) a field study in a retail store to examine effects of provision of brand-level sustainability information on choices product evaluations and retailer perceptions.

18-E: Not Green or Healthy as You Think? Misperceptions From Sustainability Labeling

Yoon-Na Cho, Villanova University, USA

The trend of environmental consciousness in food is expected to increase as consumers demand more information on environmental and social impacts. Drawing from consumers’ lay theories and the match-up hypothesis we examine the influence of healthfulness of food and sustainability levels on product evaluations to promote green food decisions.

8-I: How Temporal Separation in Budgeting Affects Spending Behavior

Yuna Choe, Texas A&M University, USA Christina Kan, Texas A&M University, USA

We propose that as the temporal separation between budgeting and purchase increases people become more willing to overspend. Bud- geting money may produce similar pain of payment to spending money and as people adapt to the cost with the the passage of time the pain associated with the decision lessens thereby increasing spending.

5-N: Take it Humorously: Effects of Humorous Rebuttals to Negative Attacks

Hsuan-Yi Chou, National Sun Yat-sen University, Taiwan Xing-Yu Chu, National Sun Yat-sen University, Taiwan Ching Chin, National Sun Yat-sen University, Taiwan

Response to election attacks might have a greater impact on the effects of the attack and on voters’ evaluations of both candidates. How- ever this issue is under-researched. This study analyzes the relative effects of different rebuttal strategies and demonstrates the moderation of the responding candidate gender and negative attack type.

6-C: Mr . Right vs . Mr . Right Now: The Impact of Male Physical Dominance on Women’s Consumption Patterns

Brett Christenson, University of Alabama, USA Christine Ringler, University of Alabama, USA Nancy Sirianni, University of Alabama, USA

Evolutionary psychology posits the motivations for consumption are made to satisfy genetic and reproductive goals. However the relationship between evolutionary drivers and gender roles has yet to be examined. We look at the interaction between male physical dominance and traditional gender roles on women’s consumption of self-enhancing versus homemaking products.

12-B: Interpersonal Face: A Major Motivation for Luxury Consumption in Asia

Jung-Han Chung, Sogang University, South Korea Young-Won Ha, Sogang University, South Korea

We investigate whether face affects consumers’ desires for luxury consumption. Across three experiments we show that interpersonal face as well as authentic pride is a critical factor in Asia and that Asian consumers are willing to pay a face premium. Marketers need to for- mulate a strategic plan to elicit face motivation for Asian consumers. 1022 / Working Papers 4-M: Psychological Antecedents of Luxury Brand Purchase Intent in Asian Market

Hoeun Chung, Fayetteville State University, USA

Despite of much attention given to the penchant for luxury consumption among Asian consumers our understanding on the underlying psychological reasons of this phenomenon is limited. The primary purpose of the current research was to explore culture-relevant psychological correlates of luxury brand purchase intent based on social consumption related theories.

8-A: Attenuating Money Illusion? How Personalization Affects the Consideration of Inflation in the Consumers’ Financial Planning

Henning Cordes, University of Muenster, Germany Carsten Erner, University of Muenster, Germany, and University of California Los Angeles, USA Thomas Langer, University of Muenster, Germany

We study how letting consumers construct their personal inflation rate affects the consideration of inflation in the long-term financial planning. We find that the personalization is not beneficial: The complex elicitation procedure reduces the subjective knowledge about inflation which induces consumers to care less about inflation in their financial planning.

9-AA: What Did You Eat at School Today? The Influence of Preschool on Children’s Food Consumption

T. Bettina Cornwell, University of Oregon, USA Sung-Hee Paik, University of Oregon, USA Eric Setten, University of Oregon, USA

This observational study shows significant differences across classrooms in the consumption of condiments vegetables grain and protein in publicly funded preschools. We theorize several possible mediators of this “classroom effect” on consumption. This research aids in the understanding of food socialization and has public policy implications for early childhood programs.

2-Y: Uncommon Beauty: Disabled Models in Ads Improve Brand Liking

Martina Cossu, Bocconi University, Italy Zachary Estes, Bocconi University, Italy

Two studies demonstrate that advertisements involving a disabled model increase brand liking. A first study demonstrates the main effect using as stimuli female disabled models. A second study replicates this result in the lab using both female and male models and tests for political correctness by measuring respondents’ response times.

18-H: The Authenticity of Corporate Social Responsibility Actions on Brand Symbolism

Diego Costa Pinto, ESPM Business School, Brazil Márcia Maurer Herter, ESPM Business School, Brazil Leonardo Nicolao, UFRGS, Brazil Mellina Terres, UFCSPA, Brazil

This research analyzes how corporate social responsibility (CSR) actions (related or unrelated to the company’s core business) influence brand symbolism. Findings suggest that CSR actions unrelated (vs related) to the company’s core business increase perception of brand social consciousness influencing brand symbolism and having downstream effects on behavioral intentions. Advances in Consumer Research (Volume 45) / 1023 11-U: Doing Good by Buying from a Peer: When and Why Consumers Prefer Peer Economy Options

John Costello, The Ohio State University, USA Rebecca Walker Reczek, The Ohio State University, USA

This research finds that consumers evaluate peer economy purchases as more prosocial than similar traditional options and that these evaluations mediate higher purchase intentions. This relative preference is attenuated when the consumer’s focus is drawn to the fact that the peer-to-peer provider is a corporate entity vs. an individual.

11-T: Carpe Diem: Examining the Impact of Hedonic Experiences on Prosocial Behavior

Daniela Cristian, BI Norwegian Business School, Norway Bob Fennis, University of Groningen, The Netherlands Luk Warlop, BI Norwegian Business School, Norway

We examine whether consumers behave more altruistically after engaging in volitional hedonic consumption. Results show that consuming pleasurable foods fosters intentional as well as factual helping behavior. This facilitating effect of hedonic consumption is more prominent for consumers that are chronically prone to disinhibit.

7-D: Do Hedonic Experiences Facilitate the Disregard of Sunk Costs?

Daniela Cristian, BI Norwegian Business School, Norway Bob Fennis, University of Groningen, The Netherlands Luk Warlop, BI Norwegian Business School, Norway

Besides enjoyment hedonic experiences lead to myopic behaviors and enhance a present-moment orientation. This seemingly “myopic” focus reduces the consideration of past and leads to positive outcomes. We show that following hedonic consumption consumers display lower sensitivity to sunk costs. Thus under particular conditions yielding to temptation is beneficial.

16-F: I’m Just Trying to Help: How Prosocial Bragging Behaviors Hurt and Help Nonprofit Organizations

Michelle Daniels, Arizona State University, USA Kirk Kristofferson, Arizona State University, USA Andrea Morales, Arizona State University, USA

Consumers frequently brag about their volunteering behaviors. While prior research has shown that bragging negatively impacts the bragger our research examines the negative spillover effects of one’s bragging on the non-profit organization. We find that viewing a volunteer’s bragging actually negatively affects both attitude toward and support for the organization.

16-I: Structure-Seeking in the Sharing Economy

Alexander Davidson, Concordia University, Canada Mohammad Reza Habibi, California State University Fullerton, USA Michel Laroche, Concordia University, Canada

The sharing economy has transformed peer-to-peer consumption into a multi-billion dollar industry. Despite its success resistance towards participation prevails and has been shown to be related to perceptions that sharing economy experiences lack structure and consistency. The current research identifies personal need for structure as a psychological barrier towards participation. 1024 / Working Papers 15-G: The Smell of Selling: Olfactory Cues as a route for Product Aesthetic Preferences

Ramona De Luca, FGV-EAESP Escola de Administração de Empresas de São Paulo, Brazil Delane Botelho, FGV-EAESP Escola de Administração de Empresas de São Paulo, Brazil

Olfactory information is not simply emotionally experienced but also cognitively processed by consumers in the marketplace. This theoretical article explores the effect of scent on decision-making. In particular we propose that scent affects stimulus processing fluency and product aesthetic preferences especially for atypical and no scent-based products.

16-J: Temporal Focus and Consumer Response to Donation Appeals

Yoshiko DeMotta, Fairleigh Dickinson University, USA

This research shows that people who are hopeful respond to a donation appeal positively when the appeal suggests their contribution enhances the wellbeing of both themselves and other people more than when the appeal suggests it enhances other people’s wellbeing. This effect reverses when people are nostalgic.

19-B: Keeping a Low Profile: When Outcome Uncertainty Promotes Behavioral Inhibition

Xun Deng, Shanghai JiaoTong University, China Liangyan Wang, Shanghai JiaoTong University, China

We propose that when waiting for uncertain outcomes of important life events people tend to display behavioral inhibition and keep a low profile in public. Three experiments support this “low-profile effect” hypothesis. The shorter the temporal distance of the outcome the stronger the low-profile effect.

20-R: Weird Products: Too Weird For Purchase, But Perfect For Sharing .

Qian (Claire) Deng, University of Alberta, Canada Paul Messinger, University of Alberta, Canada

This paper challenges the common view that weirdness is negative and usually leads to negative consumers’ responses. Specifically we found that compared to regular products consumers’ information-share (purchase) intentions and behavior toward weird products are higher (lower) because of perceived funniness (failed sense-making of the product usefulness).

2-L: Decomposing Consumer Evaluations of Brand Extensions: Market-based Congruity versus Engineering-based Congruity

Qian (Claire) Deng, University of Alberta, Canada Paul Messinger, University of Alberta, Canada

For brand extensions we identify two fundamental and robust dimensions of congruity: market-based congruity (including target-market image usage-occasion fit scale items) and engineering-based congruity (including feature function resource fit). We further find that -mar ket-based congruity has a much larger effect on consumers’ overall evaluation of brand extensions than engineering-based congruity.

10-G: Consumer Reasoning Reduces Post-Choice Satisfaction while Increasing Commitment

Alexander DePaoli, Northeastern University, USA

Past literature argues that reasoning about a product decision leads consumers to choose less satisfying options. I find that paradoxically it also increases consumers’ commitment to those options. Acting through distinct channels reasoning steers choice toward products which are rated as less satisfying while simultaneously promoting consistency with those choices. Advances in Consumer Research (Volume 45) / 1025 9-EE: Compulsive Consumption and Smoking Behavior

Timothy Dewhirst, University of Guelph, Canada Drew Harden, University of Guelph, Canada Wonkyong Beth Lee, Western University

In the spirit of transformative consumer research that aims to inform policy and improve well-being this qualitative study uses the tobacco industry’s consumer research made public from litigation to examine compulsive consumption. Compulsive consumers are linked by common traits such as low self-esteem proneness to fantasy urgency and sensation-seeking.

3-D: Exploring the Role of Consumption Values and Self-Construal in the Purchase of (In) Conspicuous Luxury Goods and Shopping Preferences: A Questionnaire Study

Karolina Dovgialo, London School of Economics, UK Ben Voyer, ESCP Europe, UK

We investigate luxury preferences for quiet vs loud luxury goods online vs offline shopping preferences and individual consumer differences. We found significant relationships between: i) bandwagon consumption interdependent self and preference for loud goods and ii) snob consumption independent self and willing to pay a premium for quiet goods.

17-M: The Dark Side of Competition: How Competition Results Predict Unethical Behavior

Rui Du, University of Hawaii, USA Qimei Chen, University of Hawaii, USA Miao Hu, University of Hawaii, USA

Our findings suggest that the association between competition results and unethical behaviors is moderated by power such that in the high power conditions losers are more likely to engage in unethical behaviors than winners. When differentiating beneficiaries of unethical behaviors a three-way interaction shows that both winners and losers lie.

20-S: When Having Too Many Facebook ‘Friends’ Hinders Your Online Social Activity: The Impact of Social Identity Complexity

Kimberly Duval, Concordia University, Canada H. Onur Bodur, Concordia University, Canada

This research shows that having more Facebook ‘friends’ decreases online engagement (likes shares comments) and purchase intentions for featured products for individuals with complex social identities. This effect is reversed however by highlighting a specific reference group when using normative messages. Findings have implications for marketers both online and offline.

9-Z: Using Practices Theories to Analyze Food Practices Related to Food Waste and Healthy Eating

Margot Dyen, Montpellier Supagro, France Lucie Sirieix, Montpellier Supagro, France Sandrine Costa, Montpellier, INRA, France

Consumers have to integrate recommended behaviors related to food waste and healthy eating. This work studies it holistically with practice theories. A two-step qualitative methodology shows that individuals and environment play variable roles in practices structuration and that each individual has various conduits of practices in his/her own framework. 1026 / Working Papers 9-II: The Resilient Consumer

Akon Ekpo, Rutgers University, USA Samantha Cross, Iowa State University, USA

This paper uses a novel mixed-method approach to investigate how behaviors and attitudes towards disability differs among disabled consumers. The authors argue that acknowledgment of one’s own marketplace vulnerabilities and societal constraints may result in greater control and awareness of the need to build and invoke resilience in the marketplace.

20-G: Is it Safe to Spread the Word? Exploring the Impact of Psychological Safety in Virtual Communities on Consumer Choice

André Escórcio Soares, Coventry University, UK Anvita Kumar, Coventry University, UK

This study introduces the concept of psychological safety as a facet of social contagion in online networks. Specifically it explores the impact of psychological safety on member participation level and subsequent influence on consumer choice. Academic and managerial implications of this study are discussed herein.

8-C: Higher Price Higher Quality? How Feeling Financially Constrained Affects Consumers’ Price-Quality Judgment

Yafeng Fan, Renmin University of China, China Jing Jiang, Renmin University of China, China

The paper examines how consumers make price-quality judgment when they feel financially constrained. The results showed that there is an interactive effect of financial constraint and value consciousness on consumer’s price-quality judgment. When feeling financially constrained consumers with high value consciousness (vs. low consciousness) tend to make higher price-quality judgment.

4-C: Culture Matters: How Power Distance Belief Influences the Effectiveness of Scarcity Appeal in Advertising

Yafeng Fan, Renmin University of China, China Jing Jiang, Renmin University of China, China

The article proposes that power distance can moderate the effect of scarcity appeal on product evaluation and the perceived threat from other consumers mediated this moderating effect. In addition we try to examine the moderating role played by product type (publicly vs. privately consumed products).

12-J: When Failure Isn’t the End: How Temporal Framing of Goal Failures Can Affect Goal Reengagement

Matthew Farmer, University of Arizona, USA Jennifer Savary, University of Arizona, USA

The way in which consumers fail goals can impact goal reengagement. We demonstrate that holding magnitude constant integration of failures into fewer (versus separation into several) incidents leads to better attitude toward and likelihood of goal reengagement. This effect reverses when failure separation scenarios are evaluated alongside failure integration scenarios. Advances in Consumer Research (Volume 45) / 1027 6-H: Marketplace Resources and Gender: A Standpoint Analysis of Status Bias, Cultural Stereotypes and the Production of Positional Inequalities

Shelagh Ferguson, University of Otago, New Zealand Jan Brace-Govan, Monash University, Australia Diane Martin, Aalto University, Finland

This study investigates the ways market resources support or hamper women negotiating a hyper-masculine context and the dynamic interactions between homosocality status gender-bias in the perpetuation of cultural stereotypes and positional inequalities. Standpoint analysis shows that highly accomplished female mountain climbers experience status gender-bias in their consumption community.

2-T: The Effect of Brand Heritage and Brand Authenticity on Consumer-Brand Relationships

Francielle Frizzo, Federal University of Parana, Brazil Jose Carlos Korelo, Federal University of Parana, Brazil Paulo Henrique Prado, Federal University of Parana, Brazil

Drawing on consumer-brand relationships theory this paper examines the brand heritage and the brand authenticity as determinants of the consumer self-reinforcing assets. The findings show that the effect of these brand stimuli on brand-self distance occurs through experiential and hedonic benefits delivered from the brand.

4-J: I Am What I Bought: A Cross-Cultural Study on Consumers’ Evaluations toward Different Types of Purchases and Advertisings

Pei-Wen Fu, National Sun Yat-sen University, Taiwan Ruoyun Lin, Leibniz-Institut für Wissensmedien, Germany Sonja Utz, Leibniz-Institut für Wissensmedien, Germany Chi-Cheng Wu, National Sun Yat-sen University, Taiwan

This study investigates how material values influence Western and Eastern consumers’ perceptions toward material and experiential purchases and whether the differences would further be reflected on their advertising attitudes. The results show that both Eastern and Western consumers have a more positive attitude toward experiential-values appeal advertisings.

8-H: Second Guess Yourself, But Do It Right: When Does Making A Second Estimate Improve Quantitative Judgments?

Celia Gaertig, University of Pennsylvania, USA Joseph Simmons, University of Pennsylvania, USA

When does averaging multiple estimates from the same person improve quantitative judgment? We find that averaging participants’ first and second estimate is less helpful when they are first asked to consider whether their first answer was too high or too low.

19-L: The Effect of Long-Term Versus Short-Term Orientation on Variety Seeking

Huachao Gao, University of Victoria, Canada Yinlong Zhang, University of Texas at San Antonio, USA

This research focuses on the understudied cultural dimension of long- versus short-term orientation. Specifically we find that long-term oriented consumers have a higher preference for dynamics and changes which in turn makes them to engage in more variety-seeking behaviors than their short-term oriented counterparts. 1028 / Working Papers 9-V: Take a Bite out of Apple: How Does Static Food Advertising Signaling Dynamic Influence Food Consumption

Fei Gao, HEC Paris, France Weiwei Zhang, University of Otago, New Zealand Tina M. Lowrey, HEC Paris, France

The present study demonstrated that static food advertisements signaling dynamic can positively influence people’s judgments toward foods and found that it influences female’s and male’s portion size choices from different directions. Furthermore we also demonstrated the role of mental imagery plays in this effect.

20-L: Sell Out Without Losing Out: Overcoming the Harmful Impact of Incentivized Reviews

Maximilian Gerrath, Leeds University, UK Bryan Usrey, University of East Anglia, UK

Despite the commonality of incentivized product reviews research widely neglected its impact on product review blogs. Three experimental examine the impact of review valence (negative vs. neutral vs. positive) disclosure (intrinsically vs. extrinsically motivated) and attachment on review credibility and blog loyalty.

10-M: Ownership Polarization: An Alternate Account of the Endowment Effect

Colleen Giblin, Carnegie Mellon University, USA Carey Morewedge, Boston University, USA

We find that merely associating an experience with the self in turn polarizes evaluations: the good things in life seem even better but the bad things seem even worse. This finding is inconsistent with previous theory supporting a new perspective on the cognitive framing effects of ownership.

5-A: Affecting Consumers: A fMRI Study on Regulatory Focus Framed Information in the Field of Animal Welfare

Nadine Gier, Heinrich-Heine-Universität, Germany Caspar Krampe, Heinrich-Heine-Universität, Germany Peter Kenning, Heinrich-Heine-Universität, Germany

Applying regulatory focus theory to animal welfare information we show that promotion focus framed information elicits greater subjective liking indicated by increased neural activity in the vmPFC. Moreover framed information influence neural processing of subsequent information demonstrated by a greater activity in ACC an effect not seen on behavioral level.

9-BB: When Innovation Collides with Nature: Mere Consideration of Novel Food Products Can Decrease Evaluations of the Entire Product Category

Justina Gineikiene, ISM University of Management and Economics, Lithuania Bob Fennis, University of Groningen, The Netherlands

We show that exposure to novel products entailing an inherent trade-off between healthiness and naturalness can lead to ambivalent evaluations. In turn such evaluations may spill over from single product exemplars and negatively affect evaluations and willingness to try the entire product category. This negative impact on ambivalence was not invariant but a function of (higher) levels of preference for predictability.

4-S: When You Expect It the Yeast: The Emergence of Craft Beer Culture in Finland

Alexei Gloukhovtsev, Aalto University, Finland

This working paper contributes to literature examining the impact of globalization on consumer behavior. Initial findings from the Finnish craft beer context illustrate how global consumer culture takes root in local hegemonic consumptionscapes by tapping into consumer resistance movements and consumer identity work with the aid of cultural intermediaries. Advances in Consumer Research (Volume 45) / 1029 10-B: Do You Feel Like a Fraud? How Experiencing the Impostor Phenomenon Influences Consumption Choices

Emily Goldsmith, Marymount Manhattan College, USA Stephen Gould, Baruch College, USA

Most people occasionally feel like a fraud. Usually we overcome these feelings by acknowledging why we deserve our accomplishments. There are times when despite all external evidence we feel like an impostor. We propose that a person experiencing the impostor phenomenon prefers products that allow them to hide fraudulent feelings.

17-O: Unaffordable Option and Sensitivity to Social Inequality

Pierrick Gomez, NEOMA Business School, France Adilson Borges, NEOMA Business School, France

This research investigates the effect of adding an unaffordable option to a choice set on sensitivity to social inequality. Two experiments demonstrate that adding an unaffordable option to a choice set fosters concern for social inequality and diminishes support for group based dominance especially for individuals high in political conservatism.

14-J: Social Comparison Direction and Its Impact on Consumer Preference for Distinctive Products

Xiushuang Gong, Renmin University, China Ying Ding, Renmin University, China Yafeng Fan, Renmin University, China

This paper examines how direction of comparison influences consumer preference for distinctive products. Across two experiments we show that consumers are more inclined to choose unique products after making upward comparisons than downward comparisons. This effect is mediated by consumer need for uniqueness.

2-Q: Low Fit in Co-Branding: Explaining the Success of Designer-For-Retailer Collaborations

Gabriel Gonzales, Pennsylvania State University, USA Johanna Slot, Pennsylvania State University, USA Margaret Meloy, Pennsylvania State University, USA

The current research finds that co-branded collaborations between low-cost retailers and high-end designers are attractive to consumers due to a ‘misfit’ between collaboration partners providing consumers with access to otherwise inaccessible products. The designer’s creative control moderates demand. Ironically product attractiveness is attenuated if prices are lower than expected.

17-P: When Not Having Enough Prompts Consumers to Show Off: Reminders of Resource Scarcity Prompt Narcissism

Laura Goodyear, Concordia University, Canada Caroline Roux, Concordia University, Canada Ali Tezer, Universite de Montréal, Canada Kelly Goldsmith, Vanderbilt University, USA

We propose that reminders of resource scarcity prompt consumers to become more narcissistic and consequently shifts their product preferences. Across three experiments we demonstrate that: scarcity (vs. control) increases narcissistic tendencies narcissism mediates the effect of scarcity on selfishness and scarcity shifts consumers’ preferences toward products with more prominent brands. 1030 / Working Papers 10-T: The Moderating Effect of Handedness on Lateral Placement Effects

Stephen Gould, Baruch College, USA Emily Goldsmith, Marymount Manhattan College, USA

This research examines differences in product preference for people based on handedness. We find that right-handers are influenced by display order while left-handers are not. No matter which side a stimulus is placed on left-handers will not vary in their choices. Right-handers favor the choice that is presented first.

2-I: Communication Style Mismatch on Twitter and Facebook: How Luxury Brands Fail to Communicate Their Desirability and Exclusivity

Simone Griesser, University of Warwick, UK Qing Wang, University of Warwick, UK Thomas T. Hills, University of Warwick, UK

The essence of luxury is exclusivity and desirability which is fostered by abstract language. Consumers use more abstract words when tweeting luxury brands than functional brands. Luxury brands responding fail to employ similarly abstract language suggesting a mismatch of communication style and hindering processing fluency which enhances brand communication evaluation.

5-F: Forget Your Regret: Invoking Post-Purchase Hyperopia Mitigates Impulse Purchase Regret

Jamie L. Grigsby, Midwestern State University, USA Robert D. Jewell, Kent State University, USA Colin Campbell, Kent State University, USA

This research investigates how regret over an impulse purchase can be reduced post-purchase thus improving customer satisfaction. If customers think about a time in the distant past in which they made a hyperopic choice they are happier with their decision to indulge in a recent impulse purchase.

7-F: Googled Experiences: Internet-induced Confidence in Decision Abilities Increases Enjoyment

Tito L. H. Grillo, University of Texas at Austin, USA Cristiane Pizzutti, Federal University of Rio Grande do Sul, Brazil Adrian F. Ward, University of Texas at Austin, USA

Consumers increasingly use the Internet to prepare for decisions. The current studies show that Internet search elicits a “feeling of already knowing” that inflates consumers’ decision confidence (S1 S2). This confidence may be self-fulfilling; searching online before choosing experiences increases consumers’ experience enjoyment independently of the option they choose (S3).

4-L: Mixed Emotions Taking Me Over: How Anticipatory Emotions and Civic Risk Determine Choices for Stigmatized Products

Jerry Grimes, Grenoble Ecole de Management, France

Fighting temptations is human nature but how exactly do we reconcile our personal desires for stigmatized products whose consumption may negatively impact our civic standing? Two studies demonstrate how consumers appraise two closely linked stigmatized products (alcohol and cannabis) differently considering situational factors one’s anticipatory emotions and civic risk perception. Advances in Consumer Research (Volume 45) / 1031 9-E: Children’s Food-related Consumer Socialization: Parental Goals and Strategies

Alice Grønhøj, Aarhus University, Denmark Malene Gram, Aalborg University, Denmark

We study food-related consumer socialization through family interviews in a qualitative study. Parents socialize children for healthy eating but other concerns such as preserving family cohesion are perceived as more important. The findings suggest that family dynamics parents’ goals strategies and context are important for understanding children’s healthy eating socialization.

9-C: Brand Authenticity Revisited: The Case of Three Times a Day

Amélie Guèvremont, Université du Québec à Montréal, Canada

This research examines the construction of an authentic brand image in an online community. Based on a netnography of Three Times a Day brand and five individual interviews four authenticity dimensions are identified: brand virtuousness brand proximity brand instinct brand imperfections. Results confirm the complex context-contingent nature of brand authenticity.

12-A: Input vs. Output-Based Framing and How They Influence Motivation

Manissa Gunadi, Erasmus University Rotterdam, The Netherlands Bram Van den Bergh, Erasmus University Rotterdam, The Netherlands

Should people be told to perform 10000 steps as fast as they can or to complete as many steps as they can in 24 hours? Our results suggest that stipulating expected output (i.e. quantity) leads to higher motivation compared to specifying expected input (i.e. duration).

17-K: Role of Group Cohesiveness in Consumers’ Responses to Mobile Promotions

Reetika Gupta, ESSEC Business School, Singapore Sourjo Mukherjee, ESSEC Business School, France Kasthuri Jayarajah, Singapore Management University, Singapore

This research examines the role of social networks in consumers’ responses to mobile promotions. Using CLT we show that in cohesive groups consumers have a more favourable attitude towards the proximal promotion (shorter lead time) than the distal promotion enabling the marketer to deploy real-time consumer-specific targeted promotion strategies.

Practicing the (un)Healthy = Tasty Intuition: Towards an Ecological View of the Relationship between Health and Taste Consumer Judgments .

Simona Haasova, University of Vienna, Austria Arnd Florack, University of Vienna, Austria Christoph Welles, University of Vienna, Austria

Do consumers apply the unhealthy=tasty intuition to real food products from a supermarket? Three studies (two representative) revealed that consumers mostly find healthier products also tastier but this relationship varies between products and individuals. Consumer judgments might rely rather on package cues simultaneously implying tastiness and healthiness than the intuition.

2-E: Are Brands Known by the Company They Keep? The Effects of Speech Rate on Employee and Brand Perceptions

Fatima Hajjat, Ithaca College, USA Elizabeth Miller, University of Massachusetts, USA

We examine the effect of employee speech rate on perceptions of employee and brand personality. Across four experiments we find that slower-speaking employees are perceived more negatively than employees speaking at a fast or normal rate. These employee perceptions then “spill over” to affect customers’ perceptions of the brand. 1032 / Working Papers 3-C: Does Incompetence Hurt More than Immorality? Construal Levels Affect Responses to Negative Brand Information

Bing Han, Shanghai Jiao Tong University, China Liangyan Wang, Shanghai Jiao Tong University, China Ke Xue, Shanghai Jiao Tong University, China

In three experiments the authors find that for consumers with a high (low) construal level incompetence (immorality) information causes lower brand evaluation than does immorality (incompetence) information respectively in manufacturing and service industries. Also we confirm that information accessibility and diagnosticity mediate this interaction effect.

2-X: To Collaborate or to Support: The Effects of Implicit Theory on Consumer Response to Anthropomorphized Brand Roles

Bing Han, Shanghai Jiao Tong University, China Liangyan Wang, Shanghai Jiao Tong University, China Ke Xue, Shanghai Jiao Tong University, China

Through three experiments in manufacturing and service industries the authors find that anthropomorphized brand roles (partner vs. servant) and consumers’ implicit theories (entity vs. incremental) can interactively impact brand evaluation as well as purchase intention. Also we confirm that consumers’ perceived efficacy mediates this interaction effect.

3-G: Reducing Regret for Maximizers: The Role of Construal Level

Louise Hassan, Bangor University, UK Edward Shiu, Bangor University, UK Miriam McGowan, Bangor University, UK

Maximizers generally regret their choices more than satisficers. We are the first to suggest a strategy easily undertaken in the marketplace to break through the regret-maximize-regret cycle. Across three experiments we show that manipulating the construal maximizers experience to more abstract level reduces the intensity of the affective response (regret).

7-L: Role of Arousal in Experiencing Feeling of Missing Out

Ceren Hayran, Doctoral Candidate in Marketing Koc University, Turkey Zeynep Gürhan-Canli, Professor of Marketing at Koc University, Turkey

This research enhances our knowledge about the consequences of feeling of missing out (FOMO) on desirable experiences that are taking place in one’s environment. Based on the circumplex model of affect we examine how the arousal dimension of a FOMO state influences individuals’ experience and product preferences in consumer context.

16-H: Secret Keeping Spurs Conformity Consumption

Dongjin He, Hong Kong Polytechnic University, China Yijie Wang, Hong Kong Polytechnic University, China Yuwei Jiang, Hong Kong Polytechnic University, China Gerald Gorn, Hong Kong Polytechnic University, China

This research shows that secret keeping increases consumers’ conformity behavior in consumption. This effect is mediated by the motivation to avoid social attention. Moreover perceived self-control moderates the effect of secret keeping on conformity such that the effect is more salient for consumers with lower perceived self-control. Advances in Consumer Research (Volume 45) / 1033 2-K: Consumer Responses to Anthropomorphized Brand Alliances

Dongjin He, Hong Kong Polytechnic University, China Fangyuan Chen, Hong Kong Polytechnic University, China Yuwei Jiang, Hong Kong Polytechnic University, China

We explore the impact of brand anthropomorphism on consumers’ perceptions and reactions toward brand alliance. In three studies we demonstrate that when a human schema was activated (vs. not) consumers felt more negatively about the termination of a brand alliance due to the relationship norm associated with humanlike alliance relationship.

13-H: The Consumption of Digital Live Content: How Live Streaming Enhances Interest in Uninteresting Content

Daniel He, Columbia University, USA Jonathan Hurwitz, Columbia University, USA Ran Kivetz, Columbia University, USA

We investigate the psychological benefits of viewing digitally live streamed content particularly content normally considered uninteresting to watch. Using an online video platform we propose and find that compared to prerecorded content live streamed content increases consumers’ intention to join in on uninteresting activities because live experiences make mundane activities feel more social.

5-D: Does Humor in Advertising Trigger Competence and Warmth Inferences about Advertisers?

Chi Hoang, BI Norwegian Business School, Norway Luk Warlop, BI Norwegian Business School, Norway Klemens Knoferle, BI Norwegian Business School, Norway

This study proposes that people make warmth and competence inferences about advertisers who express humor in advertising. Adver- tisers are perceived as competent but not warm when using incongruity to elicit laughter as both warm and competent when using arousal humor and as neither competent nor warm when using disparaging humor.

2-V: The Influence of Sex Appeal in Advertising: The Moderating Role of Brand Loyalty

Sujeong Hong, Sungkyunkwan University, South Korea

This research shows that consumers exhibiting high loyalty toward a brand react more negatively to sex appeals in brand advertising than those exhibiting low loyalty. We intended to examine an additional boundary condition under which the effect of sex appeal can be either enhanced or diminished.

13-L: When Dissimilar Others Positively Impact Online Purchase Intentions

Rahil Hosseini, Pompeu Fabra University, Spain Ana Valenzuela, Baruch College, USA & ESADE Business School, Spain Maria Galli, ESADE Business School, Spain

Research on the influence of others in product adoption has generally found a positive role of similar-other product advocates. We study the influence of dissimilar-other product advocates in online contexts and find that dissimilar advocates can encourage product adoption by catering to certain motivations particularly when perceived social presence increases.

12-I: Upward Social Comparison in Goal Pursuit: Exploring the Role of Regulatory Focus

Yu Hu, Salem State University, USA

An experiment showed that during a goal pursuit a person’s regulatory focus affected how social information is used in goal monitoring judgment. Specifically in an upward social comparison assimilation effect is more prominent in people with promotion focus whereas contrast effect is more prominent in people with prevention-focus. 1034 / Working Papers 19-K: Sleepiness, Need for Stimulation, and Variety-Seeking by Consumers

Zhongqiang (Tak) Huang, University of Hong Kong, Hong Kong Yitian (Sky) Liang, University of British Columbia, Canada Charles B. Weinberg, University of British Columbia, Canada Gerald J. Gorn, Hong Kong Polytechic University, Hong Kong

Previous research has revealed the negative impact of sleep deprivation on health but no research to date has investigated its effects on consumer behavior. This investigation fills this void by showing that sleepier consumers tend to seek more variety which is driven by a desire for sensation to maintain wakefulness.

16-P: Renting Alone? Relationship Status and Differential Appeal of Acquisition Modes

Li Huang, Hofstra University, USA Anastasiya Pocheptsova Ghosh, University of Arizona, USA

We investigate the impact of relationship status on preference for acquisition mode: renting or buying. We find that single consumers are more likely to rent products than married counterparts however buying preferences are not affected by the relationship status. This tendency is attenuated by increasing prevention focus of single consumers.

16-N: Exploring Fling Relationships with Rented Products

Li Huang, Hofstra University, USA Natalie Truong, Nanyang Technological University, Singapore Jennifer Argo, University of Alberta, Canada

How do consumers form relationships with rented products? Across three studies we found that renting a product is conceptually similar to a fling resulting in higher excitement compared to consuming an owned product. We validate the mediating role of fling perception on not only excitement but also various downstream behaviors.

20-E: How Broadcasting vs. Narrowcasting on Social Media Affects Consumer Memories

Li Huang, Hofstra University, USA Frank Zheng, University of Texas at Austin, USA Adrian Ward, University of Texas at Austin, USA

How do audiences impact the memories shared on social media? We find that sharing with a small group attenuates sharer’s memories as compared to a large group. This advantage is due to outsourcing memories to identifiable audiences and is diminished by enhancing the perceived heterogeneity of large group.

18-B: Jumping on the Bandwagon or Seeking Novelty: An Experimental Study on the Effectiveness of Heuristics to Promote Pro-Environmental Behavior.

Liselot Hudders, University of Ghent, Belgium Veroline Cauberghe, University of Ghent, Belgium Carla Mingolla, University of Ghent, Belgium

This study examines the impact of two distinct forms of heuristics (bandwagon and novelty) when promoting sustainable behavior. Results show that the effect of a bandwagon heuristic is mediated by normative influence while novelty heuristic is mediated by perceived uniqueness. In addition the self-concept regulates the susceptibility for the heuristic.

9-W: Taste Perception and Creativity

Young Eun Huh, HKUST, Hong Kong Yoonah Hong, Hongik University, South Korea Nara Youn, Hongik University, South Korea

Taste perception plays a key role in consumers’ food decisions. The current research examines consumers’ taste preference for a snack/ drink to enhance their creativity and how different tastes influence performance on a creativity task. Advances in Consumer Research (Volume 45) / 1035 17-G: How Satisfied Are You with Your Purchase?: Role of Experience Variety and Self-Construal Level

Kyoung Tae Huh, Seoul National University, South Korea Youseok Lee, Seoul National University, South Korea Sang-Hoon Kim, Seoul National University, South Korea

Prior research reveals that perceiving less variety in product usage experiences increases the satisfaction on the purchase. This research extends the finding by showing that utilitarian value plays a role in producing this effect and that this outcome is enhanced for interdependent thinkers who recognize greater similarity among product attributes.

17-J: Robo-Chat: Social Perceptions of Artificial Intelligence Assistants

Claudia Iglesias, Wilfrid Laurier University, Canada Grant Packard, Wilfrid Laurier University, Canada

We examine how the speech of artificial intelligence (AI) assistants (e.g. Siri) impacts perceptions of them as persuasion agents. A lab study reveals that first name addressing by AI assistants shifts perceptual mediators of the persuasion attempt from functional (i.e. processed like a product) to personal dimensions (like a person).

2-N: Does the Negativity Effect Really Work? The Positive-Negative Asymmetry in Computer Brand Evaluation .

Magdalena Jablonska, SWPS University of Social Sciences and Humanities, Poland Andrzej Falkowski, SWPS University of Social Sciences and Humanities, Poland

We challenge the negativity effect and present empirical evidence showing that in some situations positive information about the brand may have stronger effect on brand evaluation than the negative one. The effect is moderated by the reference point. Our findings add to the discussion on the effectiveness of negative advertising.

9-K: Healthy Cake or Decadent Cake, Could Dialecticism Be the Answer?

Alexander Jakubanecs, Norwegian School of Economics, Norway Alexander Fedorikhin, Indiana University, USA Nina Iversen, BI Norwegian Business School, Norway

Research on drivers of consumer reactions to indulgent products with healthy claims has been limited. We focus on dialectical thinking to predict and explain responses to these products within and across cultures. We find that consumers with situationally and chronically accessible dialecticism are significantly more accepting of such products.

11-E: Giving Money versus Giving Time: The Timing Effect of Thank-You Gifts on Donation Satisfaction

Miaolei (Liam) Jia, National University of Singapore, Singapore Isabel Ding, National University of Singapore, Singapore

Charity organizations often give thank-you gifts to donors. We demonstrate that when giving money receiving thank-you gifts after the donation (vs. before the donation) decreases donors’ donation satisfaction. However when giving time receiving thank-you gifts after the donation (vs. before the donation) increases donors’ donation satisfaction.

10-A: Does Purging Lead to Better Self-Control? The Role of the “Simplification” Mindset

Lei Jia, Ohio State University, USA Xiaoyan Deng, Ohio State University, USA Xiaojing Yang, University of Wisconsin - Milwaukee, USA

We propose that purging can activate a simplification mindset that features a mental process of simplifying prioritizing and abstracting. Such a mindset can carry-over to a subsequent unrelated context with decision outcomes that involve self-control. 1036 / Working Papers 14-P: The Power of Verbs: The Effect of Verbs on Purchase Intention

Miaolei (Liam) Jia, National University of Singapore, Singapore Isabel Ding, National University of Singapore, Singapore

The usage of concrete verbs such as “Grab the Gold Bars” is commonplace in marketing advertising and promotions. We demonstrate that concrete verbs (vs. adjectives) increase consumers’ purchase intention. However using explicitly instructive verbs (e.g. “Buy Now”) decreases consumers’ purchase intention.

1-L: The Spillover Effect of Beauty: Narcissists’ Overemphasis on Aesthetics in Product Choice

Lei Jia, Ohio State University, USA Xiaojing Yang, University of Wisconsin - Milwaukee, USA Huifang Mao, Iowa State University, USA Xiaoyan Deng, Ohio State University, USA

We propose that narcissism both as a personality trait and a temporary state increases the preference for aesthetics over utilitarian attributes in consumer product choice even for products with little symbolic and social signaling values and such an effect is mediated by self-possession link.

8-G: Power, Economic Situation and Financial Risk Taking

Jinfeng (Jenny) Jiao, State University of New York at Binghamton, USA Fang-Chi Lu, Korea University, South Korea

This paper focuses on the effect of an individual’s economic situation and feeling of power on financial risk taking. Across three studies we illustrate that an individual’s level of power increases his or her financial risk taking and that this effect is moderated by economic situation.

11-K: Pride and Dishonesty Why Does Authentic Pride Lead to More Dishonesty?

Jinfeng (Jenny) Jiao, State University of New York at Binghamton, USA Cathy Cole, University of Iowa, USA Gary Gaeth, University of Iowa, USA

This paper focuses on authentic pride and hubristic pride as antecedents to consumer dishonesty and has both theoretical and empirical implications. We further examine the moderating role of cognitive resources in influencing pride effects on dishonesty. We demonstrate that it is moral disengagement that mediates the found effect.

11-D: How Power States Affect Charitable Behaviors towards In-Group versus Out-group Recipients

Fei Jin, Peking University, China Ping Tu, Peking University, China Huawei Zhu, Wuhan University, China Keyi Li, Peking University, China

People primed with low power demonstrate stronger willingness to in-group than out-group members whereas individuals in high power state show similar propensities to both in-group and out-group members. For the low power helping in-group victims heightens reciprocity which in turn increases their propensity to behave generously. Advances in Consumer Research (Volume 45) / 1037 10-N: Quantification and Experience in Effortful Task: Moderating Role of Causality Orientation, Gender .

Daoyan Jin, University College of Southeast Norway, Norway Halgeir Halvari, University College of Southeast Norway, Norway Natalia Maehle, Western Norway University of Applied Sciences, Norway Christopher Niemiec, University of Rochester, USA

This study aims to conduct experiments to investigate the effect of quantification on people’s experience (e.g. vitality enjoyment) in effortful task. Another focus of this paper is to explore the moderating role of individual differences (e.g. causality orientation gender) on the relationship between quantification and effortful task experience.

14-A: An Official Apology Would End the Matter: Investigation on the Fit Between Message Content and Communication Method of Mobile Marketing (Text Message vs . Instant Message)

Sunghee Jun, Seoul National University, South Korea Kiwan Park, Seoul National University, South Korea

These days most companies are engaging in direct communication with their customers via instant messengers. However little research investigates the differences between instant messages and text messages and how marketers use them effectively. Our objective is to explore varying perceptions toward instant messages and text messages and suggest optimal strategies.

2-J: Consumer Brand Sabotage: When and Why Does it Cause Damage among Other Consumers?

Andrea Kähr, University of Bern, Switzerland Bettina Nyffenegger, University of Bern, Switzerland Harley Krohmer, University of Bern, Switzerland Wayne D. Hoyer, University of Texas at Austin, USA

This article shows that consumer brand sabotage (CBS) can damage the brand by decreasing consumers’ brand attitude and purchase intention and by increasing negative word-of-mouth. Indirect effects of CBS on consumers’ attitude and behavior through their cognitions were significant for consumers with high (vs. low) brand relationship quality only.

15-F: The Self-regulatory Power of Environmental Lighting

Seo Yoon Kang, Hongik University, Korea Nara Youn, Hongik University, Korea Heakyung Cecilia Yoon, Hongik University, Korea

The current research investigates the impacts of color temperature and its interaction with brightness on consumers’ self-control. The interaction of brightness and color temperature creates fluent (vs. disfluent) lighting condition that leads to less (vs. more) cognitively depleted mental state and consequentially enhances self-control.

11-N: The Impact of Vice Magnitude on Evaluations of Vice-Virtue Bundles

Mustafa Karatas, Koc University, Turkey Zeynep Gürhan-Canli, Koc University, Turkey

We show that low (vs. high) comparability of payoffs of vice and virtue components leads to a more abstract (vs. concrete) processing and a credentials-based (vs. credentials-based) justification process. Subsequently increasing the vice component of the vice-virtue bundle increases (vs. decreases) bundle attractiveness under low (vs. high) payoff comparability. 1038 / Working Papers 6-J: Vulnerable or Agentic? Medicalized Consumption and its Impact on Consumer Identity

Richard Kedzior, Bucknell University, USA

Over the last few decades the notions of masculinity have undergone significant changes. Traditional masculinity which is associated with the pursuit of agentic goals physical strength domination and resourcefulness had to be renegotiated to accommodate new emerging male identities. This paper investigates the impact of medicalization on consumer notions of masculinity.

2-G: Brand Activism in the Age of Resistance: What Does it Mean?

Joya A. Kemper, University of Canterbury, New Zealand Paul W. Ballantine, University of Canterbury, New Zealand

With protests becoming mainstream increased skepticism of key institutions and overall the re-emergence of populism brands have also decided to raise their voices in protest. What brand activism is and how it can be implemented is discussed.

12-H: Understanding the Motivation to Help Under Mortality Salience: Self-Worth Account Versus Cultural Standards Account

Sara Kim, University of Hong Kong, Hong Kong Youngsoo Kim, LSR/UX Lab, LG Electronics, South Korea Soraya Lambotte, AbbVie, UK Minhi Hahn, Korean Advanced Institute of Science and Technology, South Korea

Our research teases apart two accounts underlying the effect of mortality salience on prosocial behavior in prior work: self-worth enhancement and cultural standards reinforcement. We develop a framework for understanding which account primarily drives helping behavior under what circumstances and provide direct empirical evidence that these two accounts are distinct.

1-I: The Effect of Color Saturation on Satiation: High Saturation Increases Satiation of the Product

Hakkyun Kim, Sungkyunkwan University, South Korea Eunmi Jeon, Sungkyunkwan University, South Korea Jungyun Kang, National Taiwan Normal University, Taiwan

This research investigated how color saturation influences consumer behavior. In other words consumers can expect to become quickly satiated when they use products with high saturation than when they use products with low saturation of a color. Thus we demonstrated a mediating role of expected satiation between product colors and purchase intentions.

16-E: Gift Cards vs. Cash: The Effects of Money Perception on Asymmetric Preferences for Gift Cards between Gift Givers and Receivers

Min Jung Kim, Manhattan College, USA

The current research suggests that individuals will have different preferences for gift cards vs. cash that offer the same exchange value depending on their roles (i.e. givers vs. receivers) and money perceptions (i.e. whether they focus on the exchange value of money or the additional meaning of money). Advances in Consumer Research (Volume 45) / 1039 14-H: Narrative Advertising Effectiveness: The Role of Ad Relevance, Ad Vividness, and Ad Message Explicitness

Eunjin (Anna) Kim, Southern Methodist University, USA Eunseon (Penny) Kwon, Texas Christian University, USA

While prior literature has focused on the positive aspects of narrative vs. non-narrative ads we investigate a framework for narrative-advertising effectiveness. A study with 40 commercials (440 participants) confirms that ad relevance ad vividness message explicitness goal engagement emotional engagement and ad skepticism contribute to the persuasiveness of narrative advertising.

5-S: The Joint Effect of Affect and Information-processing Style on Price Judgment

Kaeun Kim, University of Massachusetts, USA Elizabeth Miller, University of Massachusetts, USA

The present research shows that affective states influence consumers’ price judgment depending on the currently accessible mode of processing. Findings from two experiments suggest that positive affect increases price-as-quality judgments only when global processing is salient whereas negative affect facilities price-quality perception when local processing is dominant.

3-A: “Mannequin and I”: How Social Distance from Mannequins and Self-Construals of Shoppers Affect Product Evaluation

Pielah Kim, Philadelphia University, USA Xiaoyan Deng, Ohio State University, USA

This research investigates the effect of mannequin design on consumer’s product evaluation. We find that interdependents but not independents perceive a closer social distance from abstract (vs. realistic) mannequins and evaluate the products wore by abstract (vs. realistic) mannequins more positively. Moreover perceived social distance mediates product evaluation.

16-G: Receiving Less for More: Inequality Aversion in Crowdfunding

Joonkyung Kim, University of Toronto, Canada

When a beneficiary has a lower economic standing than the benefactor the benefactor is more likely to give the money to the beneficiary if the beneficiary promises a low-cost token of appreciation rather than a high-cost token of appreciation. It is because the low-cost token better restores equality.

5-U: When Do People Make Affectively Negative Self-Presentation?

Dan King, University of Texas Rio Grande Valley, USA Sumitra Auschaitrakul, University of the Thai Chamber of Commerce, Thailand Zachary Mendenhall, McGill University, Canada

Our research identifies conditions under which consumers prefer negative branding –imagery and words that elicit negative emotional responses. We argue that this phenomenon is underpinned by a reflexive brain process we share with mammals called “dominance signaling” –a response that emerges when humans compete with intrasexual rivals for mating opportunities.

10-O: Sequence Fluency Effects on Consumers’ Judgments of Truth

Dan King, University of Texas Rio Grande Valley, USA Sumitra Auschaitrakul, University of the Thai Chamber of Commerce, Thailand

Brand claims come in the form of “X causes Y” in which X is the brand and Y is the purported effect. Experiments using subliminal and supraliminal manipulations show that brand claims will be judged as more truthful if they conform to symbolic sequences that are stored in the mind. 1040 / Working Papers 13-C: Creating Brand Reputation: How Employees Express Their Brand in Social Media Networks

Stephanie Kogler, University of Innsbruck, Austria Joonas Rokka, EMLYON Business School, France Andrea Hemetsberger, University of Innsbruck, Austria

This study compares the official Instagram account of the fashion brand Hollister with the private Instagram accounts of its employees. The analysis reveals that employees stabilize brand image by mimicking the brands’ visual language in terms of style postures gestures and themes.

10-K: Free to Choose: Belief in Free Will and Indecisiveness

Michail D. Kokkoris, Vienna University of Economics and Business, Austria Roy Baumeister, University of Queensland, Australia Ulrich Kühnen, Jacobs University, Germany

We explore consequences of belief in free will for consumer decision making. Studies 1 and 2 suggest that stronger free will belief is associated with reduced indecisiveness. Studies 3 and 4 show that this beneficial effect of free will belief is limited to consumers with high (vs. low) self-concept clarity.

12-G: Too Much of a Fun Thing: When Many Fun Activities Lead to Productivity

Kate Kooi, University of Miami, USA William Nowlan, University of Miami, USA Juliano Laran, University of Miami, USA

Research on goal pursuit often focuses on regulatory goals but the current research focuses on an indulgence goal—specifically having fun. We show that consumers who generate many (vs. few) means to have fun find themselves feeling productive and are therefore less likely to pursue the goal of having fun.

20-B: Investigating Online Reviews: The Moderating Role of Scarcity Appeal

Elika Kordrostami, Humboldt State University, USA Yuping Liu-Thompkins, Old Dominion University, USA Vahid Rahmani, Old Dominion University, USA

Using Amazon.com data we found that scarcity appeals make consumers pay more attention to the average rating for the item on Am- azon.com. Per the results of the OLS regression buyers on Amazon.com use the average rating of reviews as a heuristic in their purchase decision-making when scarcity appeals are present.

14-E: Do We Always Appreciate Creative Ads? Unlocking the Meaning of Highly Novel Ads

Afra Koulaei, University College of Southeast Norway, Norway

Consumers appreciate creativity but at the same time when experience uncertainty they form negative associations with novelty an attribute that makes an ad creative in the first place. This research suggests that the negative associations appraised from uncertainty can be mitigated by unlocking the meaning of the highly novel ads. Advances in Consumer Research (Volume 45) / 1041 12-E: The Impact of Purchase Motivation on Happiness: Experiential and Material Purchases

Sofia Kousi, Athens University of Economics & Business, Greece Flora Kokkinaki, Athens University of Economics & Business, Greece Stavros Drakopoulos, National and Kapodistrian University of Athens, Greece

The experience recommendation posits that experiential purchases make people happier than material purchases. The present research aims to demonstrate that this is partly due to the different underlying goals that motivate these two disposable income purchase types. To this end the results of three experiments are described.

14-R: Leaving the Lab: Can Mobile fNIRS Enhance Consumer Research?

Caspar Krampe, Heinrich-Heine-Universität, Germany Nadine Gier, Heinrich-Heine-Universität, Germany Enrique Strelow, Justus Liebig Universität, Germany Peter Kenning, Heinrich-Heine-Universität, Germany

This work aims to demonstrate that mobile fNIRS is an appropriate neuroimaging method for measuring consumers’ prefrontal neural reaction to brand related information. To achieve this we replicated the neural ‘winner-take-it-all’ effect of consumers’ first-choice-brand indicating the utility of applying mobile fNIRS in consumer research.

13-D: Extending the Herding Effect to the Consumption Experience: The Case of Online Music

Zachary Krastel, Concordia University, Canada H. Onur Bodur, Concordia University, Canada

Others’ opinions can act as a heuristic when deciding what to consume. However we show the “herding effect” also affects preferences during consumption. We find social information is more important for some individuals and for some music types and can have strong negative effects on song preferences in online environments.

16-M: Bridging the Chasm for Materialists: The Case of Renting Luxury

Goedele Krekels, HEC Lausanne, Switzerland Bruno Kocher, HEC Lausanne, Switzerland Sandor Czellar, HEC Lausanne, Switzerland Brigitte Muller, IAE de Toulon, France

Though literature would predict that materialistic people dislike renting luxury giving them less control over the product in two studies we show that for high materialists renting luxury leads to similar self-image and product perceptions as buying. For low materialists we only find positive effects for buying luxury not renting.

17-Q: When Saying Less Tells Me More: The Relation Between Socio-economic Status (SES) and Luxury Justification

Goedele Krekels, HEC Lausanne, Switzerland Sandor Czellar, HEC Lausanne, Switzerland David Dubois, INSEAD, France Gilles Laurent, INSEEC Business School, France Bruno Kocher, HEC Lausanne, Switzerland

This research examines whether luxury signalers justify their consumption and how this influences the receiver’s perceptions. We show that high SES people justify their luxury consumption less and that a lack of justification is perceived as indicating higher SES. However this can lead to negative product and personality perceptions. 1042 / Working Papers 20-F: How People Compose and Detect Insincere Product Reviews: Combining Experimental and Automated-Text-Analysis Approaches

Ann Kronrod, University of Massachusetts - Lowell, USA Jeff Lee, New York University, Shanghai, China Ivan Gordeliy, École Normale Supérieure, France

We explore how consumers compose and read true/fake reviews. We confirm predictions regarding the linguistic aspects of insincere reviews using automatic text analysis. We also find that awareness of these aspects does not improve fake review composition or detection suggesting these aspects are outcomes of deep psychological processes of lying.

2-D: A Brand Community’s Response to the Brand’s Transgression: The Role of Collective Psychological Ownership

Christina Kuchmaner, Kent State University, USA Jennifer Wiggins, Kent State University, USA Pamela Grimm, Kent State University, USA

We show that brand community members who perceive collective psychological ownership of the brand particularly those with high status within the community are less likely to engage in destructive actions in response to a brand transgression. Rather these consumers view themselves as partially responsible for helping the brand recover.

17-F: A Little Knowledge is a Dangerous Thing: When Co-Production Leads Consumers to Abandon a Firm for a Do-It-Yourself Alternative

Christina Kuchmaner, Kent State University, USA Jennifer Wiggins, Kent State University, USA Colin Campbell, Kent State University, USA

Co-production research has primarily focused on the benefits of co-production to firms. We suggest that consumers who attribute a successful co-production experience to themselves are more likely to leave the firm for a do-it-yourself alternative. Firms can mitigate this effect by manipulating the type of employee feedback given to consumers.

2-P: How Mortality Salience Impacts Consumers’ Preference for Brands

Polina Landgraf, IE University, Spain Antonios Stamatogiannakis, IE University, Spain Haiyang Yang, Johns Hopkins University, USA

We demonstrate in field and lab studies that mortality salience can dampen consumers’ preference for brands with exciting personalities but not for those with other brand personalities (e.g. sincerity). We explore potential mechanisms underlying this phenomenon. Our findings add to terror management and brand personality theories and offer managerial insights.

2-O: Environmental Threats and the Brand Incumbency Advantage

Even Lanseng, Norwegian School of Management, Norway

The idea that the incumbency advantage is limited to certain environmental threats is examined. Two studies demonstrate the incumbency advantage in the presence of pathogen threat cues and two studies show a diminished incumbency advantage in the presence of social exclusion and mating cues. Advances in Consumer Research (Volume 45) / 1043 17-N: The Dissimilarity Magnifying Bias

Negin Latifi Kasani, ESSEC Business School, France Claudiu Dimofte, San Diego State, USA

We introduce and examine a dissimilarity magnifying bias in social perception according to which others are considered first as exemplars of the least favorably perceived group to which they belong even when they are simultaneously members of more favorably perceived groups. The bias and its implications are assessed in the context of brand user group membership.

1-E: Exploring the Concept of Beauty in Consumer Research: A Multidisciplinary Framework and Research Agenda

Marina Leban, ESCP Europe, France Benjamin Voyer, ESCP Europe, France

This research lays the foundations of an interdisciplinary conceptual framework and research agenda for studying and understanding beauty consumer behavior. We integrate findings from evolutionary cognitive and cultural psychology and identify: i) the mechanisms behind beauty perceptions and ii) the role of beauty in consumers’ perceptions of products.

1-C: Differential Preference toward Aesthetic Design Elements of Airline Livery: An Examination of Design-Activity Congruency

Jeongmin Lee, Gachon University, South Korea Wujin Chu, Seoul National University, South Korea Dawon Kang, Seoul National University, South Korea Jisu Yi, Seoul National University, South Korea

Consumers seek congruency between aesthetic design and the activity being performed. As such we show that business travelers prefer airline livery designs that use darker tones primary colors and geometric lines. On the other hand vacation travelers’ design preference is split between toned-down designs and colorful designs.

10-D: “What a Loss of a Freebie!”: Promotion Framing and Consumer Product Returns

Shinhyoung Lee, Seoul National University, South Korea Youjae Yi, Seoul National University, South Korea

This research examines the effect of promotion framing on consumer product returns hypothesizing that framing a promotion as a free gift (vs. bundle no-promotion) reduces product return intention by making consumers perceive more loss in giving up a “freebie”-with-purchase.

13-E: I Want to Share My Information!: The Effects of Social Exclusion on Consumers’ Information Disclosure Intentions

Jiyoung Lee, University of Texas at Austin, USA Andrew Gershoff, University of Texas at Austin, USA

The current paper proposes that social exclusion heightens consumers’ willingness to disclose personal information to brands. Four studies demonstrate the proposed effect of social exclusion on information disclosure intentions which is driven by the desire for social connection. Two boundary conditions (i.e. brand benefit relationship likelihood) are also tested. 1044 / Working Papers 12-C: Missing, not Empty: Effects of Absence of Expected Objects in a Space on Motivation and Performance

Jaewoo Lee, Boston University, USA Kyeongheui Kim, Sungkyunkwan University, South Korea Minjung Koo, Sungkyunkwan University, South Korea Eunyoung Song, University of Florida, USA

We examine how the absence of expected objects in a space (e.g. bookshelves without books no dumbbells in the rack) impairs individuals’ general motivation to work and performance. We propose that such effects occur because exposure to spatial absence causes depletion of psychological resources or energy.

11-B: Donor Social Class Influences Appreciation of Donation

Jacob Lee, UNIST, South Korea Christine Kang, California State University Long Beach, USA Kate Min, Cornell University, USA

This research finds that charitable gift recipients appreciate gifts from lower (vs. upper) social class donors more because they incurred greater self-cost to give. We identify a boundary condition for this effect and find that the amount of the charitable gift (small vs. large) matters.

9-P: My Meal and My Recipe: Social Media Intervention in Rural Food Consumption

Eric Li, University of British Columbia, Canada Cristalle Smith, University of British Columbia, Canada Xin Zhao, Lancaster University, UK Sarah Weseen, University of British Columbia, Canada

This paper employed a social media intervention approach to examine how residents in rural communities share healthy food information through the “My Meal My Recipe” social media campaign. Our findings identify key factors to promote healthy eating in rural communities that constantly facing challenges on food security and healthy living.

4-B: Beyond Lei Feng! Global Fashion Brands and Modern Girl Archetypes in Emerging Chinese Market

Eric Li, University of British Columbia, Canada Xin Zhao, Lancaster University, UK Fiona Li, Independent Researcher, China Magnum Lam, Hong Kong Design Institute, Hong Kong Wing-sun Liu, Hong Kong Polytechnic University, Hong Kong

This paper examines the role of global fashion brands in the construction of “modern girl” archetypes in China. Through a semiotic analysis of printed advertisements published in popular fashion magazines we identified four “modern girl” archetypes and presented the dialogical relationship between brand-created imageries and local cultural imagination.

13-B: Competitive or Complementary? An Empirical Look at Interactions between Reselling and Agency Selling in Platform-Based Retailing

Qi Li, Nanjing University, China Quansheng Wang, Nanjing University, China Peijian Song, Nanjing University, China

Online platforms have become increasingly ubiquitous in retailing. We use empirical evidences to indicate the interactions between reselling and agent selling. Additionally we also recognize how the spillovers and the popularity of agent selling impact such interact relationship in a platform. Our findings provide important theoretical contributions and managerial implications. Advances in Consumer Research (Volume 45) / 1045 11-O: The Influences of Morality-Irrelevant Inspiration on the Two Faces of Morality Behavior

Jianping Liang, Sun Yat-sen University, China Zengxiang Chen, Nankai University, China Jing Lei, University of Melbourne, Australia

We propose a common antecedent for both faces of morality and found that morality-irrelevant inspirational experiences could decrease proscriptive morality and increase prescriptive morality via emotional and cognitive transcendence. This is the first paper to explore the roles played by transcendence resulted from inspiration and its consequences on source-irrelevant issues.

10-P: Solving the Paradox of a Large Assortment: The Moderating Role of Choice Mode

Mikyoung Lim, Sogang University, Republic of Korea Young-Won Ha, Sogang University, Republic of Korea

Prior research on the effect of assortment size on purchase behavior revealed that consumers prefer a large (vs. small) assortment but are less satisfied with the selected product from a large (vs. small) assortment. We demonstrate that the moderating effect of choice mode (instrumental vs. experiential) can explain this paradox.

5-T: The Silver Lining of Envy on Social Media? – The Relationships Between Post Content, Envy Type, and Purchase Intention

Ruoyun Lin, Leibniz-Institut für Wissensmedien, Germany

Three studies were conducted to examine the relationships between post content (experiential vs. material purchases) envy type (benign vs. malicious envy) and purchase intention (same vs. similar object). Results indicated that post content did not predict envy type but benign/ malicious envy predicted purchase intention of a same/similar product respectively.

13-I: The Development and Validation of a Scale for Virtual Identity versus Actual Identity

Chia-Wei Joy Lin, University of Tennessee at Chattanooga, USA Yinlong Zhang, University of Texas at San Antonio, USA

This research aimed at developing a theory-based scale for virtual identity. Five studies were conducted to examine refine and validate the proposed virtual identity scale. We demonstrated that this newly developed scale is a reliable and valid measure that captures the centrality meaning affect and behavior dimensions of virtual identity.

14-T: On the Proliferation of Latent Construct in Consumer Research: A Historical Perspective

Shan Lin, Norwegian School of Economics and Business Administration, Norway

The paper reviews constructs used in JCR papers published between 1974 and 2013 and analyzes them by network analysis. Authors visualize the state of the field together with how knowledge accumulates propose the approaches of construct development and call for attention for the usage of several problematic constructs.

20-P: Trust is Good, Control is Effort – The Paradoxical Role of Privacy Concerns in Customer Loyalty Programs

Helena M. Lischka, Heinrich-Heine-Universität Düsseldorf, Germany Verena Bergers, Heinrich-Heine-Universität Düsseldorf, Germany Peter Kenning, Heinrich-Heine-Universität Düsseldorf, Germany

Consumers’ rising concerns about data safety and privacy threats may also affect consumer behavior particularly in the context of customer loyalty card programs. Using structural equation modeling our work provides insights into the role of trustworthiness and risk beliefs investigating why consumers seldom translate their privacy concerns into actual behavior. 1046 / Working Papers 15-D: Product Touch and Consumers’ Online and Offline Buying: The Moderation Effect of Mental Representation

Wumei Liu, Lanzhou University, China Rajeev Batra, University of Michigan, USA Haizhong Wang, Sun Yat-Sen University, China

Prior research displays inconsistency on whether product touch increases consumers’ purchase intention (PI) of the product. We show that the effect of touch on PI is evident when consumers’ mental representation of a product is concrete but not when abstract. Mediators of the moderation effect of mental representation are tested.

14-C: Claiming to be Good in Parts or in Whole? The Joint Effect of Ad Headline and Body Copy on Ad Effectiveness

Xuefeng Liu, Loyola University Maryland, USA Jason Zhang, Loyola University Maryland, USA Jibo He, Wichita State University, USA

This study considers ad headline and body copy as two distinct components of an ad and shows that their effects are interactive rather than additive in terms of delivering product information. It also extends the classic hypothesis-testing framework by showing that hypotheses could be formed and tested within an ad.

7-B: Bad Experiences Increase Social Closeness Over Time

Kailuo Liu, University of Toronto, Canada Cindy Chan, University of Toronto, Canada

The paper explores how shared bad consumption experiences influence consumer relationships. Although past research shows bad experiences generate lasting personal unhappiness we find that bad experiences can benefits consumers in a new dimension – strengthening social relationships with others who shared the experiences.

2-C: “We are Breaking Up”: The Direct and Mediating Effects of Conceptual Metaphors on Brand Relationships

Alberto Lopez, Tecnológico de Monterrey, Mexico Martin Reimann, University of Arizona, USA Raquel Castaño, Tecnológico de Monterrey, Mexico

Across three experiments this research investigates the effects of relationship metaphors (e.g. “breaking-up”) on brand relationship strength. Drawing on conceptual metaphor theory this work demonstrates that consumers who break a pencil (vs. put together a puzzle) have a weaker (vs. stronger) brand relationship. These effects are mediated by persuasion.

10-J: Fluency Leads to Overconfidence in Decision Making

Fang-Chi Lu, Korea University, South Korea Jinfeng (Jenny) Jiao, State University of New York at Binghamton, USA

In this paper we explore an antecedence of overconfidence: perceptual fluency. Across our three studies we demonstrate that perceptual fluency leads to a greater overconfidence bias and that the metacognitive experience of feeling right in information processing underlies this fluency effect. Advances in Consumer Research (Volume 45) / 1047 14-D: Consumer Reactance to Persuasive Messages: How Personal Preferences Influence Attitude and Attitude Change

Sabrina Lucke, Technical University of Munich, Germany Joerg Koenigstorfer, Technical University of Munich, Germany Angela Y. Lee, Northwestern University, USA

Consumer preferences that is in our study consumers’ affiliation with the sender of a message drive not only immediate reactance to persuasion when there is a mismatch between the sender and own preferences but also a delayed increase in attitude. Negative (partly positive) emotions mediate the process.

19-M: The Impact of the Maximizing Mindset on Variety Seeking Behaviors

Jingjing Ma, Peking University, China

When consumers are making choices among a diverse portfolio of options for future consumption they tend to choose too many varieties that are inconsistent with their preferences at the time of consumption. This research shows that activating a maximizing mindset can reduce this variety seeking tendency especially for familiar products.

10-S: The Effect of Maximizing on Consumer Response to Features and Price

Zhenfeng Ma, Wilfrid Laurier University, Canada Jingjing Ma, Peking University, China

This research shows that the activation of a maximizing mindset elevates consumers’ tendency to choose feature-rich products even when these features are nonessential and these products more expensive. We argue that the reason maximizers are more feature focused is because maximizing makes consumers more promotion focused and less prevention focused.

7-E: Emotional Cocktails: Mixed Emotions, Gratifications, and the Enjoyment of Various Types of Movie Genre

Robert Madrigal, California State University at Chico, USA Colleen Bee, Oregon State University, USA Brandon Reich, University of Oregon, USA Troy Campbell, University of Oregon, USA Nathan Warren, University of Oregon, USA

This research examines the role of mixed emotions in the enjoyment of entertainment experiences. The present study shows that a mix of co-occurring opposite-valence emotions predicts overall enjoyment for films across three genres (bittersweet low-brow comedy and suspense) mediated by genre-specific gratifications.

9-HH: The Power Framework of Addictive Consumption: When, Why, and How Emotion Regulation Leads to Addictive Consumption

Olga Martin, University of Washington, USA Shailendra P. Jain, University of Washington, USA

Most explanations of addictive behaviors posit emotion regulation as the primary motivation for such behaviors. We develop a framework to understand why emotion regulation may lead to addictive consumption when this influence occurs and how it operates. We find perceived powerlessness mediates the relationship between emotion regulation and addictive consumption. 1048 / Working Papers 4-K: Methods to Improve Construct Measurement Equivalency in Cross-Cultural Consumer Research

Veronica Martin Ruiz, Iowa State University, USA Jose A. Rosa, Iowa State University, USA

The Construct Measurement Equivalency Method is a theory-grounded approach to research instrument translation in cross-cultural inquiry. It can enhance cross-cultural research validity and reliability help make target population instruments more culturally and linguistically relevant attenuate the loss of linguistic and cultural nuances and make possible a more natural response flow.

15-H: The Sophisticated Brand Scent: The Effects of Scent Type on Brand Sophistication

Márcia Maurer Herter, ESPM Business School, Brazil Diego Costa Pinto, ESPM Business School, Brazil Mellina Terres, UFCSPA, Brazil Leonardo Nicolao, UFRGS, Brazil

This research examines the effects of scent type on brand sophistication. In two studies we show that woody (vs. floral) scent enhances brand sophistication. In addition findings suggest that brand sophistication mediates the effect of scent type (woody vs. floral) on consumer’s willingness to pay.

4-D: Decomposing Cultural Effects in the Valuation of Time: Orientation, End and Progress

Yan Meng, Grenoble Ecole de Management, France Ana Valenzuela, Baruch College, CUNY, USA & ESADE, Spain

Individuals of Western cultural identity assign a higher monetary value to the future and are willing to wait a longer time for a larger amount of money than Easterners because of East Asians’ past orientation future ambiguity and belief that time-related events progress following trends that may reverse.

15-E: Putting Scent Where It Belongs: The Effects of Scent Intensity and Presentation Order on Scented Product Evaluation

Hua (Meg) Meng, Longwood University, USA Cesar Zamudio, Kent State University, USA Robert D. Jewell, Kent State University, USA

This research investigates how to effectively promote a new scented product and how to sequence its presentation alongside other scented products. This study examines the interactive effects of presentation order and scent intensity on the focal scent’s evaluation and suggests the optimal placement of the focal scent.

17-B: Every Crowd Has a Silver Lining: Threat Unleashes the Positive Aspects of Social Density

Uwe Messer, University of Bamberg, Germany Alexander Leischnig, University of Bamberg, Germany Elisabeth Distler, University of Bamberg, Germany

Does activation of the self-protection system change consumer reactions to the social store environment? Drawing from evolutionary psychology theory we show that an activated self-protection motive increases approach of socially dense retail environments. This effect is contingent on people’s general concern about interpersonal danger. Advances in Consumer Research (Volume 45) / 1049 6-D: On Biological Origins of Materialism

Uwe Messer, University of Bamberg, Germany Alexander Leischnig, University of Bamberg, Germany Elisabeth Distler, University of Bamberg, Germany

Drawing from biological research on prenatal testosterone and status drive we examine the relationship between digit ratio (a proxy of prenatal testosterone exposure) and materialism showing that higher exposure is linked to higher materialism in men. A potential reason is that men are more strongly predisposed to focus on possessions.

14-S: Liminality in Consumer Research

Laetitia Mimoun, HEC Paris, France Fleura Bardhi, City University of London, UK

This article integrates disparate consumer research on liminality a transitional state of betwixt and between. It introduces a theory of consumer liminality including a theoretical framework explaining what liminality is and what structures it and the concept of liminal consumer performances which captures how consumers do liminality.

20-H: Perseverance of the Underdog Effect in the Face of Negative Reviews

Dong-Jun Min, University of New Orleans, USA Anja Luethi, University of New Orleans, USA Kyeong Sam Min, University of New Orleans, USA

People like rooting for underdogs. However it is questionable whether their support for an underdog’s brand continues even after they hear something negative about the brand. In this research we examine when and why negative reviews are more detrimental for a top dog’s brand compared to an underdog’s brand.

1-H: The Effect of Abstract Versus Representational Art Imagery on Product Luxuriousness Perception And Evaluation

Tingting Mo, Zhongnan University of Economics and Law, China Xiaoyan Deng, Ohio State University, USA

This research demonstrates that applying abstract (vs. representational) art imagery to product design increased consumers’ perceptions of product luxuriousness and enhanced purchase intension (experiments 1 & 3). However this “luxury infusion” effect is much reduced for higher (vs. lower) value products (experiment 2).

4-O: Tension and Trust: Crossing the Border Between Hybrid Economies and Home

Michael Moorhouse, Western University, Canada June Cotte, Western University, Canada

Research into sharing (Belk 2010 2014) and hybrid economies (Bardhi and Eckhardt 2012; Scaraboto 2015) has blossomed. We outline a conceptual framework to organize theory and research on the trust required and tension inherent in sharing exchanges in one’s home.

8-B: Eliminating the Paper Trail: Consumers Prefer Using Cash over Cards for More Painful Purchases

Joshua Morris, Stanford University, USA Szu-chi Huang, Stanford University, USA

We propose that when consumers make more painful purchases (e.g. due to guilt or low transaction utility) they will be more likely to use cash than a debit or credit card in order to reduce their “paper trail” and the likelihood of recalling their painful financial activity in the future. 1050 / Working Papers 16-B: Consumer Benevolence and Risk Taking: Is Altruism Perilous?

Sudipta Mukherjee, Virginia Tech, USA Sam Bond, Georgia Tech, USA Mario Pandelaere, Virginia Tech, USA

Our research examines the relationship between consumer benevolence and risk taking. We find across three 3 studies that at both trait and situational levels benevolence is related to increased risk taking. Specifically we find that behaving benevolently results in decreased risk perception and subsequently increased risk taking.

10-I: Do Deals Really Help Save Money: Deal Salience Results in Increased Consumer Spending

Sudipta Mukherjee, Virginia Tech, USA Mario Pandelaere, Virginia Tech, USA

Consumers have the lay belief that looking for deals helps in saving money. In this paper over a series of 4 experiments we test this lay belief and find results that indicate to the contrary. We find that deal salience decreases self-control and increases consumer spending.

11-F: Looking a Free Gift in The Mouth: The Effect of Free Gifts on Charitable Behavior

Ashesh Mukherjee, McGill University, Canada Raj Raghunathan, University of Texas at Austin, USA Sumitra Auschaitrakul, University of the Thai Chamber of Commerce, Thailand

Charities often include free gifts such as cards stickers pens calendars and memory sticks in their donation appeals. Using the theoretical perspective of fit this research identifies gift origin as a new moderator of the effect of free gifts on charitable behavior.

5-H: I’m Scared, Want to Listen? Incidental Fear’s Influence on Self-Disclosure to Brands

Anupama Mukund Bharadwaj, University of Washington, USA Lea Dunn, University of Washington, USA

Self-disclosure is an increasingly popular topic within marketing yet scant literature has considered emotional impact on whether one divulges information to others. We examine the impact of fear on self-disclosure showing that people who experience a frightening event alone are more likely to self-disclose even when the act is risky.

5-E: Fight or Flight? Consumers’ Implicit and Explicit Emotional Reactions to Looming Threat

Lana Mulier, Ghent University, Belgium Hendrik Slabbinck, Ghent University, Belgium Iris Vermeir, Ghent University, Belgium

We used face reading technology to examine the emotional effects of the looming bias. We find evidence that looming stimuli elicit more negative emotions in consumers compared to receding stimuli and that animal threat leads to more negative emotions than human facial threat but only when it’s looming.

18-D: Me First, Then the Environment: Pro-Environmental Behavior in Communal vs . Agentic Narcissists

Iman Naderi, Fairfield University, USA

Across two experiments and following a new formulation of narcissism (i.e. the agency-communion model) this research investigates whether agentic and communal narcissists differ in their likelihood to engage in various forms of pro-environmental behavior. Supporting this proposition this study also shows that perceived self-sacrifice and perceived social-benefits could influence these two groups of consumers’ decision to purchase eco-friendly products. Advances in Consumer Research (Volume 45) / 1051 5-C: Do Consumers Get What They Want? How Desires Influence Consumer Happiness

Leonardo Nicolao, UFRGS, Brazil Vinicius Brei, UFRGS, Brazil

This research shows that the intensity of consumption desires predicts expected happiness for both material and experiential purchases. However this pattern does not hold for experienced happiness with past purchases. Past happiness increases with the intensity of desires for material purchases but not for experiences.

8-D: How Price Path Characteristics Shape Investment Behavior

Sven Nolte, University of Muenster, Germany Judith Schneider, University of Muenster, Germany

Consumers making financial decisions are confronted with price charts. In a laboratory experiment we strip charts of their information content by providing participants with full information via a different channel. Participants are influenced even by informationally irrelevant paths. We identify four important visual characteristics of charts that bias consumer decisions.

10-H: Creativity in Real Life: How a Creative Mindset Affects Intertemporal Preferences

Luke Nowlan, University of Miami, USA Carter Morgan, University of Miami, USA

This paper explores how thinking creatively affects consumers’ intertemporal preferences. We suggest that a creative mindset diminishes the vividness with which consumers perceive products and thus increases consumers’ patience for the product. We support this framework across two studies.

4-E: Domains of Consumer Tranquility and the “Tranquility Gap”

Hyewon Oh, University of Illinois at Urbana-Champaign, USA Cele Otnes, University of Illinois at Urbana-Champaign, USA Ravi Mehta, University of Illinois at Urbana-Champaign, USA

How do consumers leverage the marketplace to experience tranquility? We find three domains of tranquility: “everyday home” “everyday marketplace” and “extraordinary marketplace” which are shaped by the interaction between consumers’ level of control in the domain and their need to comply with gatekeepers. Specific sources of “tranquility gap” are discussed.

9-M: Improving the Efficacy of Posted Calorie Information Through the Strategic Use of Provincial Norms

Ashley Otto, Baylor University, USA Brennan Davis, California Polytechnic State University, USA Kirk Wakefield, Baylor University, USA Joshua Clarkson, University of Cincinnati, USA

This research addresses a critical issue in regard to how to improve the failed efficacy of the mandate requiring the posting of calorie information on restaurant menus. In response we propose a means by which to improve the efficacy of this mandate through the strategic use of provincial norms.

20-O: The Presence of Familiar Words Drive Social Media Likes, Comments, and Shares

Ethan Pancer, Saint Mary’s University, Canada Vincent Chandler, Saint Mary’s University, Canada Maxwell Poole, Saint Mary’s University, Canada

Consistent with a processing fluency account we find that online posts with a higher ratio of familiar words are associated with higher social media engagement. We analyzed Facebook posts and tweets from Humans of New York over a 3-year period to see how word familiarity shape these interactions. 1052 / Working Papers 14-L: The Effect of Stress on the Persuasiveness of (Central vs . Peripheral) Advertising Messages

Shirish Panchal, Wilfrid Laurier University, Canada Dr. Tripat Gill, Wilfrid Laurier University, Canada Dr. Zhenfeng Ma, Wilfrid Laurier University, Canada

The present research is the first study to examine the effect of stress on the persuasiveness of two key types of advertising messages (central vs. peripheral). Contrary to the negative connotation of stress we show the beneficial effect of moderate stress on information processing and its downstream effect.

18-G: The Asymmetric Effect of Going Green versus Going Luxury

Jooyoung Park, Peking University, China

Brand/line extension research has failed to investigate how consumers evaluate products with the same attributes differently depending on the focal attribute of their parent brand. This study shows that a pro-environmental product extended by a luxury brand is better accepted than a luxury product extended by a pro-environmental brand.

10-L: Number-location Bias: Do Consumers Correctly Process the Number?

Jihye Park, Hankuk University of Foreign Studies, South Korea Yoon Jin Ma, Illinois State University, USA

This study investigated whether consumers process numeric information with spatial cues and whether a number-location bias is attenuated or strengthened by verticality of visual focus. Results of a series of experiments revealed that the verticality effect was stronger than the horizontality effect of numeric information on perceived magnitude.

6-I: The Impact of Mixed-Sex Competition on Women’s Choice of Androgynous Products

Jooyoung Park, Peking University, China Fayrouz Souissi, Peking University, China

The simultaneous emergence of unisex dressing with women entering competitive environments makes it judicious to study the popularity of androgynous products in the context of mixed-sex competition. This study aims to fill a gap in the literature by investigating women’s choice of androgynous products in the context of mixed-sex competition.

11-G: Pride and Donation Behaviors

Sungjun (Steven) Park, Korea Advanced Institute of Science and Technology (KAIST), South Korea Yong J. Hyun, Korea Advanced Institute of Science and Technology (KAIST), South Korea

By using the donation amount of real money results indicate that participants experiencing authentic pride placed a more lenient bar to them while requiring others a stricter moral bar. The results become insignificant to participants when experiencing hubristic pride.

14-F: Free But Suspicious Gift Promotion: the Influence of Product-gift Fit on Devaluating the Promoted Product

Yookyung Park, Seoul National University, Korea Youjae Yi, Seoul National University, Korea

The present research demonstrates that when a low-fit gift (vs. a high-fit gift) is provided upon purchase consumers’ judgments on the promoted product’s value are undermined. This influence is mediated by perceived deceptiveness of the offer and moderated by consumer shopping orientation. Advances in Consumer Research (Volume 45) / 1053 2-S: The Asymmetric Attribution toward Brands (Underdog vs . Top-dog) upon Brand Crisis (External vs . Internal)

Kiwan Park, Seoul National University, Republic of Korea Yaeri Kim, Seoul National University, Republic of Korea Seojin stacey Lee, Seoul National University, Republic of Korea

We identify that in service domain where perceived warmth from the brand is particularly important service failure especially caused by the underdog brand is considered more serious. However in product context where brands’ competence is more important than warmth people express less forgiveness intention on top-dog brands’ crisis.

1-J: The Effect of Incidental Disgust on Aesthetic Preference

Kiwan Park, Seoul National University, South Korea Seojin Stacey Lee, Seoul National University, South Korea Joonkyung Kim, University of Toronto, Canada

Current work shows that an exposure to disease related stimuli will lead to a more favorable evaluation for simple designs and less favorable evaluation for designs visually complex. This research contributes to the emotion and art literature by showing how avoidance tendency triggered by disgust can influence aesthetic evaluation.

19-H: An Individual Choice vs . A Bundle Choice: The Effect of Choice Mechanics on Diversification Bias

Jihye Park, Hankuk University of Foreign Studies, South Korea

To the extension of Mittelman et al. (2014) this study examined how choice mechanics influence variety seeking in the multiple product choice context. The expected satiety drove variety seeking in the individual choice mechanics whereas relationships of items were more likely to be considered when consumers chose a bundle.

13-F: Location-based Services Effect on Online Purchase Intention

Jean-Eric Pelet, ESCE International Business School, France Jashim Khan, University of Surrey, UK Kelly Cowart, University of South Florida Sarasota-Manatee, USA

This paper presents an empirically tested model explaining the relationship between social media location-based services (LBS) user emotion and online purchase intention.

14-G: Incorporating Vanity into a Luxury Value-Attitude-Behavior Model- Evidence from Luxury Restaurant Consumers

Norman Peng, University of Westminster, UK Annie Chen, University of West London, UK Li-Wei Mai, University of Westminster, UK Stella Kladou, Sheffield Hallam University, UK

Existing theories of the effect of luxury value on consumers may require adjustment when applied to service-based products. To contribute to the literature this study incorporates a “vanity” variable into a luxury value-attitude-behavior model. A total of 150 participants from the UK completed questionnaires. All hypotheses are supported. 1054 / Working Papers 3-B: Can Implicit Theory Influence Construal Level?

Sara Penner, University of Manitoba, Canada Olya Bullard, University of Winnipeg, Canada Kelley J. Main, University of Manitoba, Canada

This research examines the relationship between implicit theory and construal level. Three studies show that holding an incremental (vs. entity) theory is associated with abstract (vs. concrete) information processing. The likely underlying mechanism is positivity shift that results from holding an implicit (vs. entity) theory.

16-L: When is More Better? Will How Many People Ask Affect the Outcome?

Sara Penner, University of Manitoba, Canada Kelley Main, University of Manitoba, Canada Jennifer Argo, University of Alberta, Canada

Why do many non-profits use multiple solicitors in their donation requests with the belief it will improve their fundraising results? We explore this through the lens of attitude and discover that men and women have very different reactions related to the gender and number of people asking for the gift.

11-M: Puritan Peers or Egoistic Entrepreneurs? An Examination of Moral Identity in Collaborative Consumption

Rebeca Perren, California State University San Marcos, USA Kristin Stewart, California State University San Marcos, USA

Despite proponents of collaborative consumption portraying peers as moral citizens of society recent findings suggest that egoistic motives drive participation. Platform-providing firms rely on users’ cooperative behaviors; thus this research examines how prolonged participation diminishes moral identity. Findings reveal important implications for the success of emerging peer exchange business models.

18-F: Redeeming Guilty Pleasures - Sustainable Consumption Across Vice and Virtue Categories

Anne Odile Peschel, Aarhus University, Denmark Jessica Aschemann-Witzel, Aarhus University, Denmark

The trend of upcycling is observable in different areas of consumption – including in the food industry as ‘waste to food’. While consumers avoided upcycling in virtue categories it seemed to offset the guilty pleasure of consuming vice products. Therefore it could be an alternative approach for marketing sustainable ingredients.

9-Y: Too Large to Eat it All: How Package Size Impacts Anticipated Food Waste, Intentions and Preferences

Olivia Petit, INSEEC Business School, France Renaud Lunardo, KEDGE Business School, France Bradley J. Rickard, Cornell University, USA

Through three experiments we demonstrate that larger packages prompt anticipated food waste which then reduces purchase intentions. In addition results reveal that when being primed with food waste consumers are less likely to prefer large packages even if such packages are associated with a promotion. Advances in Consumer Research (Volume 45) / 1055 9-U: Smell-o-Vision: Olfactory and Visual Store Atmospherics Prime Healthy Food Choices

Megan Phillips, Auckland University of Technology, New Zealand Sommer Kapitan, Auckland University of Technology, New Zealand

The present research investigates the ability of store atmospherics to nudge shoppers towards purchasing healthier foods. An observation study and an experimental study together demonstrate that store atmospherics can be tailored to create a message of healthfulness and act as a prime to persuade shoppers to purchase healthier foods.

5-I: Judgement and Decision-Making: The effect of induced emotional uncertainty on predicted utility and forecasting accuracy.

Athanasios Polyportis, Athens University of Economics & Business, Greece Flora Kokkinaki, Athens University of Economics & Business, Greece

The purpose of the research is to explore the effects of the certainty-uncertainty appraisal dimension of incidental emotions on predicted utility and affective forecasting accuracy. Emotions associated with uncertainty seem to result in smaller forecasting error and utility overpre- diction. To this end the results of two experiments are described.

6-F: The Red Derogation Effect

Nicolas Pontes, Queensland University of Technology, Australia JoAndrea Hoegg, University of British Columbia, Canada

This research demonstrates that the red-attractiveness effect of women viewing men is moderated by mate retention goals and that the effect is mediated by perceived threat. Additionally we show that the red-threat link is moderated by self-control strength such that the effect dissipates when self-control resources have been depleted.

11-H: Being Good Versus Being the Better: Consumer Responses to Preferential Treatment .

Vivian Pontes, Queensland University of Technology, Australia Nicolas Pontes, Queensland University of Technology, Australia Dominique Greer, Queensland University of Technology, Australia

This research shows that extent to which preferential treatment can lead to negative versus positive emotions. In particular when a benefit given causes some harm to others the positive effects of being advantaged are mitigated by the arousal of negative moral emotions which in turn decreasing customer satisfaction.

5-K: Pride and Shame Moderate the Effect of Regulatory Focus on RiskTaking Propensity

Aylar Pour Mohammad, Bocconi University, Italy Gulen Sarial Abi, Bocconi University, Italy

How self-conscious emotions affect risk taking behavior? Prior literature has indicated that promotion (prevention) focus is associated with risk seeking (risk avoiding) propensity (Bryant and Dunford 2008). Two experiments demonstrate that positive (negative) self-conscious emotions may attenuate the effect of promotion (prevention) focus on risk taking propensity.

7-J: Making Money Meaningful: Fostering Meaningful Lives Through Growth-Based Experiential Purchases

Emily Powell, New York University, USA Adam Alter, New York University, USA

As consumers strive for “the good life” they must search for both happiness and meaning. Across two studies we show that consumers who spend more money on growth-based experiences (as opposed to purely enjoyment-based experiences) derive greater meaning from those experiences and in their lives as a whole. 1056 / Working Papers 20-Q: Understanding Word-of-Mouth Transmission

Ashleigh E. Powell, RMIT University, Australia Adrian R. Camilleri, RMIT University, Australia Angela R. Dobele, RMIT University, Australia Constantino Stavros, RMIT University, Australia

We integrate emergent consumer psychology findings regarding emotional arousal self-enhancement and communication context by examining how these factors interact to influence word-of-mouth transmission.

17-L: Social TV and Viewing Experience

Emily Powell, New York University, USA Alixandra Barasch, New York University, USA

We examine the effect of social TV on viewers’ likelihood to watch a show again. We find that social TV increases repeat viewing when viewers are alone (but not in a group) and when they create content related to the show (but not unrelated content).

9-S: Presentation Style Affects Decision Processes: A Dietary Choice Study

Mona Prakash, Duke University, USA Jonathan Winkle, Duke University, USA Nicolette Sullivan, Duke University, USA Gavan Fitzsimons, Duke University, USA John Pearson, Duke University, USA Scott Huettel, Duke University, USA

When making decisions the available options can often take a variety of forms. For example menus can have some options that have images while others are described in words. This paper explores how the visual form in which an option is presented can bias decision processing.

11-Q: The Viciousness and Caring of Sharing: Morality and Motivations of Online Shamers

Chen Pundak, Tel Aviv University, Israel Yael Steinhart, Tel Aviv University, Israel Jacob Goldenberg, Interdisciplinary Center Herzliya, Israel

We focus on public shaming in online social networks. Shaming entails two contradictory outcomes: informal enforcement against deviant behavior and a violation of privacy rights. A set of studies shows the duality of shaming and that identification of the wrongdoer moderates the effect of morality on participation in public shaming.

15-B: It ‘Sounds’ Healthy to Me! The Influence of Sound on Food Choice.

Eduardo Rech, UFRGS, Brazil Cristiane Pizzutti, UFRGS, Brazil Eric Yorkston, Texas Christian University, USA

Studies on sensory stimulation has given little attention to sound influencing consumers’ choice. One of our purposes is to contribute with transformative research by exploring a way to increase consumption of healthy food. Our first results are encouraging since it is showing significant associations between sound and choice for food. Advances in Consumer Research (Volume 45) / 1057 7-I: Is a Smartwatch a Fashion Item or an IT Device?: Moderating Effect of Category Frame in Smartwatch Adoption

Jong-Youn Rha, Seoul National University, Korea Jin-Myong Lee, Chungnam National University, Korea Sunny Park, Seoul National University, Korea Eunsun Cho, Seoul National University, Korea Bo Han Lee, Seoul National University, Korea

A smartwatch is an innovative product that has both IT and fashion product attributes. Based on previous studies that suggest consumers develop knowledge schema for new products using prior knowledge of existing product categories we investigated the moderating effect of the category frame on consumers’ intention to adopt smartwatches.

10-U: To Search or Not to Search? Exploring Consumer’s Sensitivity to Task Importance

Yefim Roth, Israel Institute of Technology, Israel Kinneret Teodorescu, Israel Institute of Technology, Israel

Previous search literature suggests that consumers “do not search much”. We show that in certain settings consumers search extensively. Furthermore they often do not allocate their search effort effectively searching equal amount of time or even longer in situations with minor or even negligible benefit for their effort.

9-DD: “Works With/For You!” How Framing Health-Related Products as Partners Versus Servants Impacts the Consumption of Indulgences

Caroline Roux, Concordia University, Canada Kelly Goldsmith, Vanderbilt University, USA Kamila Sobol, Concordia University, Canada Laura Goodyear, Concordia University, Canada

We investigate the impact of engaging with a health-related product that is framed as either a partner (co-producer of the benefit) or servant (provider of the benefit) on goal-conflicting behavior. Specifically we find that framing a health-related product as a servant (vs. partner) increases enjoyment from and desire for indulgences.

2-M: Distancing from the Envied Person: How Envy, Personal Distance and Framing Brand Availability Affects Brand Evaluations

Rajat Roy, Bond University, Australia Subimal Chatterjee, State University of New York at Binghamton, USA

Feeling malicious (relative to benign) envy engenders more distancing from the envied person making a brand that is not preferred by the envied person more attractive than the brand that is and this effect is exaggerated if the brands are in limited supply rather than in excess demand.

9-F: Destigmatizing the Overweight Consumers: A Communal Perspective

Nada Sayarh, University of Geneva, Switzerland

While stigma is socially detrimental consumer researchers have little to say about how it relates to consumption. Also this concept has been mainly investigated from an individual perspective. In this paper we indicate how consumption communities have the potential to transform the stigmatized reduce their stigma and improve their well-being. 1058 / Working Papers 14-N: The Effect of Public Commitment in Consumers’ Satisfaction

Fernanda Scherer, UFRGS, Brazil Cristiane Pizzutti, UFRGS, Brazil Clara Koetz, Rennes School of Business, France

Our study demonstrates that public commitment with an imminent consumption experience (e.g. check-in in social media) increases customers’ expectations which negatively impacts on their subsequent satisfaction with the product. This mediation of the expectation showed to be total when the performance was high but partial when the performance was low.

11-A: “Doing Good and Having Fun” - The Role of Moral Obligation and Perceived Enjoyment For Explaining Foodsharing Intention

Nadine Schreiner, Heinrich Heine University Düsseldorf, Germany Sarah Blümle, Heinrich Heine University Düsseldorf, Germany Peter Kenning, Heinrich Heine University Düsseldorf, Germany

This research focus on a combination of morality and hedonism to explain foodsharing behavior. A modified version of field theory was applied to differentiate the consumer’s and provider’s perspective. Results confirmed that foodsharing can be explained by moral obligation and perceived enjoyment even though the latter is comparatively weak.

19-E: Surprise me! How Uncertainty Labels Affect Product Consumption

Anika Schumacher, Maastricht University, The Netherlands Caroline Goukens, Maastricht University, The Netherlands Kelly Geyskens, Maastricht University, The Netherlands Martin Reimann, University of Arizona, USA

Uncertainty labels such as “mystery” or “surprise” are frequently used to encourage product choice e.g. “surprise sale” Kinder Sur- prise®. Yet the actual consumption consequences of these labels are poorly understood. Across five studies we show that labels of uncertainty increase indulgent consumption holding the actual level of uncertainty constant.

18-C: Let Market Know How Your Firm is Implementing Corporate Social Responsibility

Junhee Seok, Seoul National University, Korea Sarang Go, Seoul National University, Korea Youseok Lee, Seoul National University, Korea Byungdo Kim, Seoul National University, Korea

So far many studies examined the effect of CSR on firm value but the results have been contradictory. According to our findings publi- cizing CSR is the key variable that affects firm value and the corporate reputation mediate this relationship. Furthermore the effect of CSR varies depending on the firm size.

4-R: Transnational Consumption

Zahra Sharifonnasabi, University of London, UK Fleura Bardhi, University of London, UK

We examine transnational lifestyle to identify the nature of consumption when consumers inhabit multiple countries simultaneously. A distinct notion of home emerged: fragmented transnational home anchored in various localities. We conceptualize transnational consumption as fragmented and localized where specific consumption goals are appointed and practices are associated in each locality. Advances in Consumer Research (Volume 45) / 1059 2-Z: When Not to Surprise? The Role of Rewards and Communal Mindset on Perceptions of Brand Value

Sirajul Shibly, State University of New York at Binghamton, USA Subimal Chatterjee, State University of New York at Binghamton, USA

We show that surprise relative to certain rewards positively impact brand-value perceptions among consumers who see the rewards as a signal of building a communal relationship. Surprise hurts value perceptions of consumers with less communal mindset particularly for cash rewards (relative to gift cards).

14-B: Can’t Switch Off: The Impact of an Attentional Bias on Attitudes

Sunaina Shrivastava, University of Iowa, USA Gaurav Jain, Rensselaer Polytechnic Institute, USA Dhananjay Nayakankuppam, University of Iowa, USA Gary Gaeth, University of Iowa, USA

We show that individuals’ attention gets chained in incomplete events even when that attention is normatively unwarranted i.e when the outcomes are known and no further action is needed. We show that such a non-normative chaining of attention leads to negative attitudes towards entities associated with the event.

17-H: Publicity as Justification

Kao Si, The Chinese University of Hong Kong, China Xianchi Dai, The Chinese University of Hong Kong, China

Whereas public displays of negative conducts are undesirable we show a counter-intuitive positive effect of publicity on people’s inference and evaluation of negative conducts. Observers evaluate an apparently negative behavior more favorably when it is done publicly than privately. Evidence supports an inference mechanism of the current effect.

19-G: Hot Streak! Consumer Inferences from Streaks of Virtuous Choices

Jackie Silverman, University of Pennsylvania, USA Alixandra Barasch, New York University, USA Deborah Small, University of Pennsylvania, USA

In three studies we demonstrate that people infer that others are more likely to choose a virtuous option over a vice option after a recent streak compared to other patterns of choices even when holding the base rate of choosing virtue constant. This effect is driven by higher perceived commitment to the virtuous behavior.

11-P: The Positive Effect of Moral Licensing on Human Behavior: When a Virtuous Act Motivates You to Work Hard

Anirban Som, IIM Trichy, India

Moral licensing theory argues that individuals who initially exhibit moral behaviors subsequently resort to immoral behaviors. Literature in the context of moral licensing predominantly highlights the negative consequences of this phenomenon. The current research contributes to this literature by highlighting the positive effects of moral licensing on individuals’ behaviors.

1-D: Effects of Logo Colorfulness on Consumer Judgments

Jiaqi Song, Hong Kong Polytechnic University, Hong Kong Yuwei Jiang, Hong Kong Polytechnic University, Hong Kong Gerald J. Gorn, Hong Kong Polytechnic University, Hong Kong

Consumers have more favorable attitudes toward a brand when its logo contains multiple colors than a single color. This effect is driven by an increase of the perceived product variety of the company and moderated by brand positioning and the external attribution of the logo colorfulness. 1060 / Working Papers 9-Q: Packaging Glossiness Thwarts Perceived Food Healthiness

Jiaqi Song, Hong Kong Polytechnic University, Hong Kong Yuwei Jiang, Hong Kong Polytechnic University, Hong Kong Anish Nagpal, University of Melbourne, Australia Gerald J. Gorn, Hong Kong Polytechnic University, Hong Kong

Two studies demonstrate that consumers judge food and beverages with glossy exterior packages as less healthy than those in matte packages. This effect is driven by the belief that food and beverages in glossy packaging are more processed and moderated by the stereotyped perceived healthiness of the food category.

1-K: The Impact of Visual Glossiness on Psychological Newness and Product Judgment

Jiaqi Song, Hong Kong Polytechnic University, Hong Kong Yuwei Jiang, Hong Kong Polytechnic University, Hong Kong Gerald J. Gorn, Hong Kong Polytechnic University, Hong Kong

Two studies in this research show that consumers have more favorable evaluations of products with a glossy as opposed to a matte surface. In addition this effect is found to be driven by the heightened psychological newness of products with a glossy appearance.

7-N: “You Grin At Each Other, You Celebrate Together, You Hug Strangers”: Consequences of Shared Identity Within Crowds at Sporting and Music Events

Maximilian Stieler, University of Bayreuth, Germany Fergus G. Neville, University of St. Andrews, UK Claas Christian Germelmann, University of Bayreuth, Germany

Building on contemporary crowd psychology this paper investigates if and how individuals experience a shared identity with strangers at diverse crowd events. We conducted qualitative semi-structured interviews with crowd participants in three different crowd contexts. Find- ings suggest that shared identity led to increased intimacy with strangers and intense positive feelings.

3-E: Neurological Evidence for an Interrelation Between Imagery, Psychological Distance, and Construal

Paul E. Stillman, Ohio State University, USA Hyojin Lee, San Jose State University, USA Xiaoyan Deng, Ohio State University, USA Rao Unnava, University of California Davis, USA Kentaro Fujita, Ohio State University, USA

While many researchers have investigated the effects of visualization in marketing limited research has directly explored the nature of mental images due to methodological limitations. In the present research we employ fMRI technique to investigate consumers’ mental imagery. We found a fundamental relationship between mental imagery psychological distance and construal.

16-O: I Am What I Borrow: Trait Transfer in Borrowed and Rented Products

Rusty Stough, University of Wisconsin, USA Page Moreau, University of Wisconsin, USA

We extend work on identity by examining how consumers view themselves through products they either rent or borrow. We found an increase in satisfaction with rented products from an anonymous source than from borrowed products and evidence for trait transfer when products are borrowed. Advances in Consumer Research (Volume 45) / 1061 5-L: Psychological Drivers of Customer Compliance: Self-Identity Threats and Discrete Emotions

Kristina Stuhler, Old Dominion University, USA Chuanyi Tang, Old Dominion University, USA Mahesh Gopinath, Old Dominion University, USA

This research develops and tests a theoretical framework on consumer compliance by incorporating self-identity theory and cognitive-emotion theory. We propose self-identity threats and discrete emotions influence customer compliance. A qualitative study will be conducted first to develop and refine the theoretical framework which will then be tested by experimental design.

10-F: Choice of Ownership – How Choosing Enhances the Endowment Effect on Product Pricing and Evaluation

Daniel Sun, University of Calgary, Canada Mehdi Mourali, University of Calgary, Canada

We all have possessions but not all of our possessions are valued equally. The initial choice of ownership may trigger subsequent differential connections to a particular possession. This research investigates the effect of choice on product ownership in terms of both pricing and evaluation.

19-F: Vegas it Up to Boost Your Performance!

Arash Talebi, ESSEC Business School, France Sonja Prokopec, ESSEC Business School, France

Two pilots and one lab study show that using uncertain price promotions leads to desirable placebo effects. This research expands the scope of marketing placebo effects to desirable placebo effects in promotional settings. Moreover we propose that the effect is mediated by enhanced feelings of empowerment due to heightened self-attribution.

4-A: Beauty is in the Eye of the Beholder: Young Girl’s Perception of Female Attractiveness in New Zealand, India, and Fiji

Tabitha Thomas, University of Otago, New Zealand Kirsten Robertson, University of Otago, New Zealand Maree Thyne, University of Otago, New Zealand

This research investigates 6-12-year-old girls’ perception of female attractiveness in three countries. The findings underscore the importance of understanding how the beauty ideals and stereotypes prevailing in the society influences young girls’ notion of what constitutes beauty and the impact it can have on their well-being.

4-I: Good Looks, Possessions, and Stereotypes: The Relationship Between Physical Attractiveness and Materialism

Tabitha Thomas, University of Otago, New Zealand Kirsten Robertson, University of Otago, New Zealand Maree Thyne, University of Otago, New Zealand

Prior research has shown that internalizing materialism and appearance related ideals negatively impact adults’ well-being however little is known whether young children also internalize similar ideals. Thus in the present research using Dittmar’s (2008) framework we examine the ages at which children start perceiving links between physical attractiveness and materialism. 1062 / Working Papers

9-N: Medicalization in Finnish Functional Foods: The Normalization of Medical Disorders

Jack Tillotson, Aalto University, Finland

I examine the normalization of medical disorders as consumers organize around discourses surrounding the functional food market. Medicalization – the definition and treatment of social problems as medical issues - is a widespread phenomenon but understudied in consumer research. This paper looks at functional foods as a form of medicalization.

2-F: Are the Eyes the Mirror to the Soul?: The Influence of Eye Gaze Direction on Narrative Transportation and Self-Brand Connection

Ngoc (Rita) To, University of Houston, USA Vanessa Patrick, University of Houston, USA

We aim to demonstrate the influence of ad model’s gaze direction (direct versus averted) on consumer self-brand connection. With two studies we demonstrate that averted (versus direct) gaze enhances narrative transportation which drives self-brand connection. We also examine the moderating role of rational (vs. emotional) appeals on this relationship.

7-A: Assembling the Ritual: Consumer Experiences in a Diverse Network

Omer Torlak, Turkish Competition Authority, Turkey Mujdat Ozmen, Eskisehir Osmangazi University, Turkey Muhammet Ali Tiltay, Eskisehir Osmangazi University, Turkey Mahmut Sami Islek, Anadolu University, Turkey Ufuk Ay, KTO Karatay University, Turkey

We investigate the formation of consumer’s consumption ritual parts and discourses associated with Feast of Sacrifice. This study contributes to the literature on religious rituals and practices because of viewing ritual as an assemblage including material and expressive features as well as human and non-human actors.

15-A: Does it Sound Crowded to You? A Proposed Multisensorial Approach to Crowding Research

Fernanda Trindade Deyl, Pontifícia Universidade Católica do Rio Grande do Sul, Brazil Lélis Balestrin Espartel, Pontifícia Universidade Católica do Rio Grande do Sul, Brazil

Research regarding crowding is broad in academia; however few studies focus on its auditory perspective. The present study proposes a broader approach to crowding research by putting emphasis on the auditory variables present in retail environments which may contribute to the perception of density and consequent perception of crowding.

14-K: The Effect of Service Providers’ Attractiveness on Consumers’ Preferences for Status Quo

Natalie Truong, Nanyang Technological University, China Jennifer Argo, University of Alberta, Canada George Christopoulos, Nanyang Technological University, China

How does beauty influence the consumer’s decision process and choices? In two studies we found that with exposure to attractive (vs. unattractive) female salespersons female participants were more likely to prefer the traditional over the innovative products (study 1) and demonstrated higher preference for status quo (study 2). Advances in Consumer Research (Volume 45) / 1063 5-W: Why Firms Should Be Thankful for Thankful Customers: Incidental Gratitude Increases Satisfaction with Consumption Experiences

Arianna Uhalde, University of Southern California, USA Valerie Folkes, University of Southern California, USA

Three studies demonstrate incidental gratitude influences consumers’ reactions to product failure experiences. Incidental gratitude (vs. non-emotionally infused thoughts about others incidental pride in oneself and incidental admiration of others) 1) increases satisfaction by fostering compassion and 2) reduces behaviors like spreading negative WOM and grudge-holding against the firm.

4-P: The Emancipatory Alliance for Social Activism: The Symbiotic Interplay between Subcultures and Social Movements

Emre Ulusoy, Youngstown State University, USA

While subcultures mold potentially destructive youth rebellion into constructive activism and mobilize individuals to social movements social movements imbue subcultures with a radical political edge the two working together to seek meaningful social and cultural change.

20-D: False Positives: How Double-sided Review Systems Affect Consumers’ Rating Valence

Bryan Usrey, University of East Anglia, UK Antje Graul, Leeds University, UK

Drawing on reciprocity literature three studies demonstrate that double-sided review systems in peer-to-peer service schemes positively affect consumers’ rating valence due to activating their psychological perception of “pre-ciprocity” of the review – leading to false positives. The effect is strengthened in a socially distant and attenuated in a private review scenario.

10-Q: The Compromise Effect in Post-Purchase Consumption Behavior: Evidences from Field Experiments .

Veronica Valli, University of Mannheim, Germany Florian Stahl, University of Mannheim, Germany Elisa Montaguti, University of Bologna, Italy

This paper investigates the impact of the compromise effect (Simonson 1989) on post-purchase consumption. Results of a field experiment suggest that the selection of an intermediate option vs. an extreme one in a choice set leads to an increase in the number of (complementary) items purchased after the first choice.

17-E: Glass Houses: Disability Perceptions in the Consumption Environment

Helen van der Sluis, Arizona State University, USA Adriana Samper, Arizona State University, USA Kirk Kristofferson, Arizona State University, USA

While marketing and public policy decision-makers express heightened sensitivity to stigmatized groups such as people with disabilities little research has examined their impact on the consumption environment. We find that people view disabled individuals higher in moral character but also punish them more harshly when they engage in moral violations. 1064 / Working Papers 9-D: Bring Back My Childhood! Nostalgia as Mediator of Retro-Food and Consumer Based Brand Equity

Caty Velez, Tecnológico de Monterrey, Mexico Miriam C. Flores, Tecnológico de Monterrey, Mexico

This research will expand the field/area of retro-branding by recognizing the effect of nostalgia as well as authenticity in the relationship between a retro-food product and Consumer Based Brand Equity (CBBE). A mixed methodology that includes a netnography and two quantitative studies will be used.

11-W: The Impact of Perceived Temperature on Responses to PSA ads

Akshaya Vijayalakshmi, Indian Institute of Management Ahmedabad, India Meng-Hsien (Jenny) Lin, California State University Monterey Bay, USA Melika Kordrostami, California State University San Bernadino, USA

This working paper finds that the affiliative sensations triggered through priming of warm/cold touch can increase the effectiveness of public service announcement (PSA) by increasing empathy threat perceptions and donations to the cause. However this is likely to be true only for high need for touch participants.

19-I: How Brand Preference Similarity Influences Married Couples’ Brand Variety Seeking Over Time

Anna Vredeveld, Berry College, USA Selcan Kara, University of Massachusetts - Dartmouth, USA

This research examines how married consumers’ brand preference similarity influences brand variety seeking behavior. Findings contribute to extant work on shared brand consumption and variety seeking by illustrating how romantic relationship partners engage in purpose- ful brand variety seeking as a part of their relationships.

12-D: The Impact of Brand Exposure on Perceived Goal Progress and Goal Satiation

Darlene Walsh, Concordia University, Canada Francis Jarry, Concordia University, Canada

Though prior research finds that exposure to a brand can activate brand-related goals we show that brand exposure can satiate brand-related goals. Further we show that goal satiation occurs because consumers who are susceptible to goal satiation also experience perceived progress towards achieving the brand-related goal.

9-I: Eating with Friends: Preference for Handmade Foods

Xin Wang, Nanjing University, China Chunqu Xiao, Nanjing University, China Hong Zhu, Nanjing University, China

What people choose to eat varies as a function of relationships with co-eater(s). When eating with friends consumers prefer handmade foods relative to manufactured foods. They prefer manufactured foods when with new acquaintance. When they try to strengthen social relations they choose handmade foods.

17-C: Fluid Compensation: The Role of the Interdependent Self

Qin Wang, Arizona State University, USA Monika Lisjak, Arizona State University, USA Naomi Mandel, Arizona State University, USA

When consumers experience a threat to the self they may cope by engaging in fluid compensation—affirming the self in important domains that are unrelated to the threat. Three experiments show that people with an interdependent (vs. independent) self are more likely to engage in fluid compensation to offset self-threat. Advances in Consumer Research (Volume 45) / 1065 4-F: Essentialism Increases Status Consumption of High-Class (not Low-Class) Consumers

Xue Wang, Chinese University of Hong Kong, China Ying-Yi Hong, Chinese University of Hong Kong, China Robert S. Wyer, Chinese University of Hong Kong, China

We proposed that essentialist beliefs of social class (i.e. social class is immutable and biological-based) would increase status consumption of high- but not low-class people. With both measuring and manipulating essentialism four studies consistently supported the hypothesis. We further found that entitlement could explain the relationship.

19-J: Show Me More! Powerlessness Drives Variety Seeking

Wangshuai Wang, Shanghai Jiao Tong University, China

This research investigates the impact of power on variety seeking behavior. Results from four studies provide convergent support that feeling powerless relative to feeling powerful nudges individuals toward variety seeking. Additionally we reveal that perceived autonomy mediates this relationship.

5-G: How Notifications of Product Changes Influence Consumer Ambivalence

Yiru Wang, Kent State University, USA Jennifer Wiggins, Kent State University, USA César Zamudio, Kent State University, USA

Notifications of product changes could increase consumer ambivalence towards the product. We find that this will be mitigated when the notification is framed to be consistent with the timing of the notification. Specifically when the change is in the distal (proximal) future the notification should be framed abstractly (concretely).

10-V: Truncated Text, Truncated Thought

Adrian Ward, University of Texas at Austin, USA Noah Castelo, Columbia University, USA Kurt Gray, University of North Carolina, USA

We show that frequency of text messaging is negatively associated with the ability to understand logic and meaning in context. Fre- quently engaging in the kind of truncated thought characteristic of text messaging appears to cause a similarly truncated style of thinking.

19-C: Keeping the Streak Alive!

Danny Weathers, Clemson University, USA T. Andrew Poehlman, Clemson University, USA

Consider a man who has eaten a Big Mac daily since 1972 or a runner who’s run a mile 2034 straight days. What motivates people to engage in ‘streaks’? Our research suggests streaks serve to motivate people in identity relevant domains and temporal specificity is a cue for continuance.

9-B: A Nuanced Nutritional System Facilitates the Recognition of Healthy Options, Increases Sales and Choice of Healthy Foods: A Comparison Between 3-Colors and 5-Colors Traffic-Lights Systems

Carolina O.C. Werle, Grenoble Ecole de Management, France Kévin Roche, Grenoble Ecole de Management, France Olivier Trendel, Grenoble Ecole de Management, France Amanda Yamim, Grenoble Ecole de Management, France

Simplified nutritional-labeling systems provide single indicators of nutritional quality. Four studies compared two nutrition-labeling systems varying in complexity. A more subtle system (traffic-light with five colors instead of three) increases healthy choice and purchase. The 5-color system facilitates nutritional information processing and this influences healthiness perception of food products. 1066 / Working Papers Moderately Simple Fbdg Work Better than Complex or Oversimplified Recommendations to Influence Food Consumption in a Real Setting

Carolina O.C. Werle, Grenoble Ecole de Management, France Kévin Roche, Grenoble Ecole de Management, France Olivier Corneille, Université Catholique de Louvain, Belgique Caroline Roux, Grenoble Ecole de Management, France

Food based dietary guidelines (FBDG) are regularly updated by governments but their efficacy is questioned. Previous research suggested that simple actionable guidelines can change behavior but evidence confirming their efficacy is still scarce. Patrons of a restaurant received tray-sets containing FBDG varying in complexity. Moderately simple FBDG influenced food consumption.

2-B: “Reclaiming the News”: Episodes of Contention in the News Field

Verena E. Wieser, University of Innsbruck, Austria

This study adopts a field-theoretical lens to investigate how incumbents and challengers negotiate power relations in the news field in the digital age. Interviews with market experts in German-speaking countries reveal how traditional newspaper brands navigate episodes of contention through collective attribution of threats and organizational appropriation.

4-N: Perspective Flexibility Matters: The Influence of Cultural Syndromes on Attribution of Service Failures

Vincent Chi Wong, Lingnan University, Hong Kong, China Feifei Huang, Chinese University of Hong Kong, China

Customers with a collectivistic (vs. individualistic) orientation or a long-term (vs. short-term) orientation are likely to attribute a service failure more to service provider’s contextual factors and less to service provider’s dispositional factors. These effects are mediated by the flexibility of perspectives customers take when making a judgment.

14-Q: Trend versus End State: The Role of Implicit Theories in Interpreting Communications Conveying a Partial Reduction in Negative Product Attributes

Vincent Chi Wong, Lingnan University, Hong Kong, China Lei Su, Hong Kong Baptist University, China Howard Pong-Yuen Lam, Chinese University of Hong Kong, China

The results of four experiments show that the communication of partially reduced negativity of a product attribute will lead to higher (lower) product evaluations among consumers who are incremental (entity) theorists. This communication effect is mediated by consumers’ trend-based versus end-state-based interpretation of the reduced-negativity communication.

6-B: How Scarcity Influences Mate Preference in Romantic Relationship and Subsequent Mating Strategy?

Yinghao Wu, Renmin University, China Jing Jiang, Renmin University, China

The current research proposes that perceived general scarcity which includes but far beyond a scarce mate pool will polarize people’s mate preference for romantic partners that such selectivity differs across gender and mate-value and that those with relatively low mate value will adopt collective strategies to enhance their attractiveness. Advances in Consumer Research (Volume 45) / 1067 5-J: Partner or Servant: How Anthropomorphized Brand Role Releases the Negative Effect of Social Exclusion

Yinghao Wu, Renmin University, China Jing Jiang, Renmin University, China

The current research shows that socially excluded (vs. included) consumers are willing to pay more for anthropomorphized brands. More importantly when brand plays a servant role people with high (vs. low) self-esteem are likely to pay more but such a discrepancy disappears when brand plays a partner role. The need for control recovery mediates the above relationship.

16-K: The Impact of CSR Transgressions on Consumer Support for Non-profits: The Role of Negative Moral Emotions and Political Identity

Chunyan Xie, Western Norway University of Applied Sciences, Norway Richard P. Bagozzi, University of Michigan, USA Silvia Mari, University of Milano-Bicocca, Italy

We develop and test new psychological mechanisms underlying consumer support for nonprofits as a function of perception of CSR transgressions. Results showed that the moral emotion of contempt mediates the relationship between perceived corporate community transgressions and consumer support. Further political identity moderates the effects of transgressions on contempt.

11-L: Prosocial Consequences of Corporate Social Irresponsible Actions: The Role of Negative Moral Emotion, Moral Identity, and Empathy

Chunyan Xie, Western Norway University of Applied Sciences, Norway Richard P. Bagozzi, University of Michigan, USA Silvia Mari, University of Milano-Bicocca, Italy

We add to extant CSR research by addressing prosocial consequences of corporate social irresponsible actions and providing a psychological mechanism underlying such a novel effect. Results showed that contempt mediates the relationship between corporate community transgressions and consumer support for nonprofits. Further moral identity and empathy moderate the effects.

9-G: Diets with Cheat Days: The Effect of Implicit Self-Theories on Recovery from Goal-Inconsistent Behavior

Yi Xie, Arizona State University, USA Naomi Mandel, Arizona State University, USA Meryl Gardner, University of Delaware, USA

Dieters fall into two types: Abstainers who completely avoid temptations and moderators who occasionally break the rules. We show that lay beliefs about the renewability of self-control determine which approach people choose. Our results suggest that there is no “best” strategy: dieters are more successful when they follow their beliefs.

9-H: Does Larger Portion Sizes => Lower Calories? How Healthiness Beliefs Impact Calorie Estimation

Jieru Xie, Virginia Tech, USA Fengyan Cai, Shanghai Jiao Tong University, China Rajesh Bagchi, Virginia Tech, USA

We demonstrate a paradoxical relationship between portion size and caloric content where a larger- (vs. smaller-) sized healthy food item is judged as containing lower calories. This occurs because of a misguided notion that if a food-item is healthy then more of it is healthier. 1068 / Working Papers 5-R: The Feeling of Being Moved

Ji (Jill) Xiong, National University of Singapore, Singapore Leonard Lee, National University of Singapore, Singapore Yih Hwai Lee, National University of Singapore, Singapore

We propose and demonstrate that being moved is a mixed emotion that connects with self-transcendence core value(s) of perceived personal importance. Thus marketing communications such as advertisements that elicit the feeling of being moved increase brand affiliation by highlighting the core value that consumers share with the focal brand.

6-G: Wearing V Neck, Getting More Trust: An Evolutionary Psychology Approach to Examine the Effect of Collar Style on Trust

Jialiang Xu, University of Manitoba, Canada Fang Wan, University of Manitoba, Canada Chenbo Zhong, University of Toronto, Canada

“What a strange power there is in clothing” ---- Isaac Bashevis Singer. Adopting evolutionary psychology perspectives we conducted two experiments and found that one’s sweater Collar styles (V neck vs turtleneck) can affect others’ trust of him and interaction with him.

11-R: Two Faces of Moral Identity Internalization: The Interactive Effect of Lighting, Moral Identity Internalization and Recipient’s Cost on Prosocial Behavior

Jun Yan, University of Manitoba, Canada Fang Wan, University of Manitoba, Canada Luke Zhu, University of Manitoba, Canada Aaron Stone, University of Manitoba, Canada

This paper tests an interactive effect of lighting and moral identity on prosocial behavior. Individuals low in Moral Identity Internaliza- tion are more likely to help when the surrounding is bright and the recipient’s cost is low. In contrast individuals high in MII help regardless of lighting and the recipient’s cost.

5-V: When Sadness Comes Alive, Will It Be Less Painful? The Effects of Anthropomorphism on Sadness Regulation

Li Yang, Tsinghua University, China Fangyuan Chen, Hong Kong Polytechnic University, Hong Kong Rocky Peng Chen, Hong Kong Baptist University, Hong Kong

Can anthropomorphism contribute to the regulation of sadness? In three studies we demonstrated that anthropomorphic (vs. neutral) thinking decreased the intensity of sadness experience. We suggest that psychological detachment explains this effect. This result offers fresh insights to research on emotions and anthropomorphism and has implications for consumer well-being.

6-E: The Effects of Different Parenting Cues on Consumer’s Regulatory Focus: A Parental Investment Perspective

Chun-Ming Yang, Ming Chuan University, Taiwan

With two studies this research examined the idea that cues emphasizing different aspects of parental investment could lead to different regulatory focus. Specifically cues about prevention (promotion) parental investment (e.g. providing foods) can activate prevention (promotion) focus. We also observed a regulatory fit effect in a marketing communication context. Advances in Consumer Research (Volume 45) / 1069 16-C: Craving for Hope: Scarcity Cues and Their Influences on Persuasiveness of Hope-Appeal Advertising

Chun-Ming Yang, Ming Chuan University, Taiwan Tzu-Yun Yang, Ming Chuan University, Taiwan Shu-Ni Hsu, National Taiwan University of Science and Techonology, Taiwan

Three studies examined the idea that the salience of resource scarcity decreases consumers’ prosocial behaviors (e.g. intention to donate and willingness to pay for fair-trade products) and this effect is mediated by consumers’ sense of control. The authors also argue that the presence of hope appeal may reverse this relationship.

11-V: Roughness Increases Prosocial Behavior When People Believe the World Is Fair

Ning Ye, Temple University, USA Maureen Morrin, Temple University, USA

Our research shows that people with higher belief in a just world (BJW) are more likely to help after being exposed to rough (vs. smooth) haptic image. However for people with lower BJW roughness makes no difference in the helping intention.

15-C: Now It’s Personal: The Influence of Touch on the Construction of Preferences

John Yi, University of Arizona, USA Jesper Nielsen, University of Arizona, USA

Using a tablet versus a mouse when shopping online has shown to increase a shopper’s preference for the chosen item. Further study demonstrates that the physical motions relevant to forming a preference increase this preference effect and it is not the preexisting preferences that drive the effect.

11-C: Feature the Benefactor or the Victim? How Charity Advertisements with Different Protagonist Foci Affect Donation Behavior

Bingqing(Miranda) Yin, University of Kansas, USA Jin Seok Pyone, University of Kansas, USA Surendra Singh, University of Kansas, USA

We examine charity appeals with different protagonist foci and demonstrate that a charity appeal featuring an identified benefactor can promote more charitable donation compared with a comparable victim-focused charity appeal. Results are explored via moral elevation and charity attitude resulting from portraying a benefactor.

16-D: Don’t Appreciate Your Customer Too Much: Business Gift Giving and Verbal Acknowledgement

Yanfen You, University of Wisconsin - Milwaukee, USA Massimiliano Ostinelli, University of Wisconsin - Milwaukee, USA Xiaojing Yang, University of Wisconsin - Milwaukee, USA

We examine how verbal acknowledgment of customer value could make business gifts backfire. Business gifts increase customer attitude when customer value is not salient. However their effects could be diminished or become negative when marketers highly acknowledge customer value. 1070 / Working Papers 20-I: Political Ideology and Review Evaluation

Sik Chuen Yu, University of Sydney, Australia Donnel Briley, University of Sydney, Australia Pennie Frow, University of Sydney, Australia

This research proposes that political ideology affects reviewer similarity’s influence on review evaluation. Those who are increasingly conservative tend to have stronger relational motivations. Consequently conservatives value reviews from similar reviewers more than dissimilar ones while liberals do not value reviews from similar and dissimilar reviewers.

20-A: A Taste of Discrimination: A Study on Consumer Reviews on Yelp

Chao Yu, Cornell University, USA Drew Margolin, Cornell University, USA Xian Wang, Tsinghua University, China

This paper reveals an asymmetric-discriminative pattern in consumers’ online evaluation-making mechanism. We examine how consumers from different social classes discriminate in evaluations on Yelp. A multi-level random-effect model shows discrimination exits in consumers’ favorable reviews to outgroups but this effect does not hold for negative evaluations.

13-A: Cyber-Heterotopias in Social Media Platforms: The Illusion of Intimacy in Ephemeral “Stories”

Mariella C. Zavala, University of California Irvine, USA

Using Foucault’s notion of heterotopia (1986) this paper conceptualizes the “Stories” feature found in some social media platforms as cyber-heterotopias. Their ephemeral quality and an illusion of intimacy are presented as defining of their heterotopic distinction in an effort to contribute to our understanding of the consumer relations they affect.

20-C: Advice-Seeking Intention: The Interplay of Regulatory Focus and Problem Type

Xianfang Zeng, University of Calgary, Canada Mehdi Mourali, University of Calgary, Canada

This research investigates how regulatory focus and problem type jointly affect advice-seeking intention. When faced with factual problems individuals with a prevention (vs. promotion) focus tend to seek accuracy-based advice; in contrast when presented with judgmental problems those with a promotion (vs. prevention) focus tend to solicit autonomy-based advice.

2-W: The Unofficial Name: Effects of Brand and Product Nicknames on Brand Perception

Zhe Zhang, University of Houston, USA Vanessa Patrick, University of Houston, USA

We explore the effects of brand nickname (vs. formal name) use on brand perception. We show nicknames anthropomorphize the brand and transform the consumer-brand relationship from transactional to relational. This influences consumers’ self-brand connection and brand attachment which leads to enhanced willingness to purchase additional products from the brand.

2-H: Brand Hate

Chun Zhang, Concordia University, Canada Michel Laroche, Concordia University, Canada

The grounded theory is applied to study the construct brand hate. Seventy-one hated brands are reported through twenty-five in-depth interviews. Three facets of associations of brand hate are presented: emotional cognitive and physical. Four facets of reasons are discovered and six different levels of consequences are obtained. Advances in Consumer Research (Volume 45) / 1071 20-N: The Effect of Uncertain Rewards on Customer Co-promotion Intention: An Exploration of Optimistic Bias

Yuan Zhang, Xiamen University, China Zhaoyang Guo, Xiamen University, China Yirang Zhang, Xiamen University, China

Three studies suggest that uncertain rewards affect customers’ co-promotion intentions. Study 1 revealed that consumers receiving a lower uncertain reward showed more sharing intention of promotion information and the mechanism was optimistic bias. The controllability facilitated the uncertainty effect (Study 2) and the delay reward inhibited the effect (Study 3).

9-J: Familiarity and Affiliation with Anglo Names Affect Purchase Decisions of Chinese Food

Xian Zhao, University of Kansas, USA

For an unfamiliar Chinese dish attaching it to an Anglicized Chinese name increased Americans’ willingness to pay than attaching to an original Chinese name; for the familiar Chinese dish attaching it to an original Chinese name increased anti-multiculturalist Americans’ willingness to pay than attaching to an Anglicized Chinese name.

20-M: The Dynamic Diverting Effects of User-generated Contents vs. Firm-generated Contents on Product Browse

Yingnan Zhao, Nanjing University, China Xin Wang, Nanjing University, China Quansheng Wang, Nanjing University, China Peijian Song, Nanjing University, China

The research aims to identify the effects of visits to user-generated and firm-generated contents on visits to product in E-commerce. These effects are different in short- and long-term for new and existing users. Vector auto-regression model indicates that there’s difference between the effects of visits toward UGC and FGC.

9-X: The Devil is in the Smile: Smiley Temptation and Motivated Calorie Estimation in Restrained Eaters

Xiaoying Zheng, Nankai University, China

Using smileys in tempting food is a popular product design strategy in the marketplace. However little has been known about how consumers perceive such smiley temptations. Across two studies this research finds that restrained eaters perceive a smiley temptation as containing more calories than a non-smiley temptation.

17-I: Relational Greeting: When and When not to Treat Your Consumers as Family

Qichao Zhu, Tsinghua University, China Maggie Wenjing Liu, Tsinghua University, China Chuang Wei, Tsinghua University, China

This research explores effects of relational greeting on consumer purchase intentions. Compared to stranger greeting contact employees’ relational greeting towards consumers can lead to higher purchase intentions when consumers’ product knowledge is high while the reverse happens with low knowledge. Consumers’ feeling of control mediates the effect of relational greeting. 1072 / Working Papers 10-R: The Effect of Free Shipping on Consumer Patience

Ozum Zor, Rutgers University, USA Rafay Siddiqui, Hong Kong Polytechic University, Hong Kong Ashwani Monga, Rutgers University, USA

For standard (low fee; slow delivery) versus expedited (high fee; fast delivery) shipping we examine the dynamic that arises when standard shipping is free (vs. low fee). We show an interaction effect: making standard shipping free (vs. low fee) decreases sensitivity to changes in the fee charged for expedited shipping. DataBlitz

Maximizing Utility but Minimizing Wellbeing: A Meta-Analysis on the Maximizers’ Paradox

Alex Belli, University of Technology Sydney, Australia François Carrillat, University of Technology Sydney, Australia

In an attempt to settle the debate on the relationship between maximization and well-being a sub-group meta-analysis was conducted accounting for well-being dimension valence (positive or negative). Findings revealed maximization is detrimental to well-being regardless of the dimension valence but is moderated by the maximization scale and by long-term orientation.

Mental Stimulation and Cognitive Orientation: Implications for Children’s and Adults’ Food Choices

Dipayan Biswas, University of South Florida, USA Annika Abell, University of South Florida, USA Courtney Szocs, Portland State University, USA

The results of two field experiments with children at a middle school cafeteria and two lab experiments with adults show that undertak- ing mental stimulation tasks (such as by solving math problems) before making a food choice leads to less indulgent food choices.

There Is No Such Thing as a Free Lunch: The Negative Effect of Zero Price on Consumer Demand

Fengyan Cai, Shanghai Jiao Tong University, China Xiaomeng Fan, Northwestern University, USA

Both intuition and empirical evidences indicate that people like cheap products and free products are particularly attractive. But we showed a backfiring effect of zero price on consumer demand and identified a moderator that could drive the effect of zero price on demand into either positive or negative direction.

The More Elongated, the More High-end: The Effect of Package Shape on Perceived Brand Status

Huan Chen, Renmin University of China, China Jun Pang, Renmin University of China, China Minkyung Koo, University of Illinois at Urbana Champaign, USA

This research examines the effect of package shape on consumers’ perceived brand status. Results from three studies show that consumers perceive a product with tall & slim (vs. short & stout) shape to have a high brand status. Moreover perceived attractiveness is the mechanism underlying this effect on brand status perception.

What Make Products Weird? Antecedents to and Underlying Mechanism of Weirdness .

Qian (Claire) Deng, University of Alberta, Canada Paul Messinger, University of Alberta, Canada

Little research attention is paid to the concept of weirdness because of the negative association of weirdness. This paper challenges this view not only by identifying the key antecedent (extreme incongruity) to the weirdness and the underlying mechanism (failed sense-making) but also by showing its marketing potential (generating more WOM).

Isabel Ding, National University of Singapore, Singapore Leonard Lee, National University of Singapore, Singapore

Consumers often receive feedback in goal pursuit. This article examines the effect of feedback frequency on consumers’ goal performance — does higher frequency feedback lead to higher goal performance? Four studies demonstrate that while consumers hold the lay belief that higher (vs. lower) frequency feedback leads to higher goal performance empirical results show that it leads to lower goal performance. This occurs because higher frequency feedback increases the level of perceived goal progress which reduces consumers’ tendency to continue pursuing the goal. This effect only holds when the feedback information is relevant to goal progress but is attenuated when the information is irrelevant. Furthermore this effect is attenuated when the goal is specific.

The More You Know the More You Search: Post-Decision Information Search and the Effect of Prior Knowledge and Maximizing Tendencies

Maura Ferreira, Pontifical Catholic University of Rio Grande do Sul, Brazil Cristiane Pizzutti, Federal University of Rio Grande do Sul, Brazil

Previous literature argues that cognitive dissonance and feelings of regret or dissatisfaction increase the amount of information sought in post-decision timespan. We show that the knowledge individuals hold before the decision-making point predicts post-decision information search as well. Additionally we demonstrate that high levels of maximizing tendencies boost this effect.

The Motivating Effect of Expectation-Inconsistent Social Information

Katherine Flaschen, Stanford University, USA Szu-chi Huang, Stanford University, USA

We explore the impact of expectations about social others on goal-directed behavior. Across four studies we find that people experience a motivational boost from witnessing another person’s expectation-inconsistent (versus expectation-consistent) behavior but only when such behavior is goal facilitative (e.g. observing an unfit person jogging).

Yuck! Feelings of Disgust, Self-threats and Compensatory Behaviors .

Elena Fumagalli, HEC Paris, France L.J. Shrum, HEC Paris, France

Consumers frequently encounter disgusting images. We examine how physical and moral disgust differentially affect consumers’ identity and compensatory consumption. We show that physical disgust decreases consumers’ sense of power which prompts them to consume conspicuous goods. In contrast moral disgust decreases consumers’ self-esteem causing them to act prosocially.

The Effect of Information About Previous Donation Impact on Recurring Donations

Zohar Gilad, Technion – Israel Institute of Technology, Israel Liat Levontin, Technion – Israel Institute of Technology, Israel

Nonprofits experience around 90% donor attrition rates. Our aim is to understand the antecedents of recurring donations and develop strategies to cope with attrition. As donors are seldom informed about their donations’ impact we suggest and find that informing donors about the positive impact of their donation increases subsequent donations.

The Psychophysiology of Touch-Sensitive Interfaces: Somatosensory Encoding of Intensity, Pleasantness, and Technology-Induced Affect-Regulation

Christian Hildebrand, University of Geneva, Switzerland

Building on the segregation of affective touch in the somatosensory cortex we hypothesize and show that touch-sensitive interfaces cause a substantial increase in consumers’ arousal relative to valence perceptions (using both objective and subjective measures of affect) ultimately promoting a greater impulse to purchase and willingness-to-pay for a focal product. Advances in Consumer Research (Volume 45) / 1075 Choosing None versus Choosing “None of the Above”: The Effect of No-Choice Decision Expression on Choice Outcomes

Seung Eun (Sonia) Kim, Seoul National University, South Korea Sung J. Jung, INSEAD, Singapore Jin M. Kim, Yale University, USA Kyoungmi Lee, Seoul National University, South Korea

When a ballot does not provide a selectable “none of the above” option voters will cast blank ballots if none of the candidates are desirable – or will they? In this research we investigate whether the choice context in which people must express their no-choice decision alters their choice outcomes.

Friends Without Benefits? How the Costs of Having a Relationship Influence Risk-taking

Jennifer K. Lee, University of Southern California, USA Lisa A. Cavanaugh, University of British Columbia, Canada

While having close relationships is known to increase risk-taking we identify conditions leading to reduced risk-taking. When consumers focus on the costs (rather than benefits) of having a relationship and when the risk consequence affects both relationship partners risk-taking decreases. However when individuals lack these close relationships risk-taking behavior increases.

Inspired to Speak Up: Role of Inspiration on Minority Opinion Expression

Xingbo Li, University of Louisville, USA Zengxiang Chen, Sun Yat-sen University, China

Expressing unpopular or minority opinions has consequences even though the opinion holders may have their groups’ best interest at heart. The current research examines when and why inspiration prompts people to express minority opinions.

Effects of Implicit Theories on Customers’ Satisfaction with Service Recovery: The Value of Process

Xiaoyan Liu, Southwestern University of Finance and Economics, China Shaobo Li, Nanyang Technological University, Singapore Elison Lim, Nanyang Technological University, Singapore

Three studies examine how implicit theories impact customers’ satisfaction with service recovery. We find that compared to entity theorists incremental theorists express higher satisfaction and are more likely to revisit the firm when the recovery is high in interactional justice regardless of whether or not an appealing compensation is provided.

Incentive Design with Uncertainty: Keep the Hope Alive

Xiyueyao Luo, Chinese University of Hong Kong, China Luxi Shen, Chinese University of Hong Kong, China

Shen et al. (2015) shows consumers work harder for uncertain rewards than for certain rewards. How can marketers design an effort-score-payoff system with uncertainty? We find that to boost effort investment within in a time limit adding uncertainty to the score-payoff part is more effective than the effort-score part.

When Compensatory Consumption Backfires: The Pain and Pleasure of Experiential Purchases

Zichuan Mo, Peking University, China Jingjing Ma, Peking University, China

Prior research has shown that self-threats can evoke consumption that signals accomplishments in the domain of the threat. However we show that this within-domain compensation is likely to backfire especially when consumers compensate with experiential rather than material purchases. Identity relevance of experiential purchases could be the mechanism. 1076 / Working Papers Appetite for Destruction: Attractive Faces Alter People’s Food Choices

Tobias Otterbring, Aarhus University, Denmark, & Karlstad University, Sweden

Directly contradictory to the predictions made by marketing professors exposure to attractive (vs. unattractive) opposite-sex faces made people choose unhealthy rather than healthy foods. This effect primarily influenced people who rated themselves as inferior on self-view-relevant attributes but did not generalize to attractive (vs. unattractive) same-sex faces.

Emoji and Brand Engagement on Social Media

Ethan Pancer, Saint Mary’s University, Canada Lindsay McShane, Carleton University, Canada Maxwell Poole, Saint Mary’s University, Canada

Brands both human and corporate are increasingly communicating with their social media audience using emoji. The current work examines whether emoji use shapes online engagement (likes & retweets). Two field studies demonstrate that emoji presence increases engagement with brands. Facial emoji were particularly effective at increasing engagement with organizations.

The Effect of Competitive Rivalry Associations on Brand Extensions

Nicolas Pontes, Queensland University of Technology, Australia Vivian Pontes, Queensland University of Technology, Australia

This research demonstrates the extent brands benefits from existing competitive rivalry associations. In particular it shows that market leaders (but not challenger brands) benefit from a second mover advantage even when facing established well-known brands in the extension category. Further we show that a low category fit hinders these effects.

Feeling Flattered or Flawed? The Effects of Consumer Embarrassment on Self-perception

Tracy Rank-Christman, University of Wisconsin - Milwaukee, USA Danielle Warren, Rutgers University, USA

In a sequence of studies we find that embarrassment stemming from personal mistakes (negative trigger) and special treatment (positive trigger) has a negative impact on consumers’ self-perceptions. These effects are augmented by emotional contagion and mediated by feelings of social rejection.

Triangulating the Breadth of Explanatory Depth

Scott Roeder, Washington University in St. Louis, USA Leif Nelson, University of California, Berkeley, USA

We argue that the illusion of explanatory depth is attenuated not only by explanations of the focal item itself but also by explanations of other entirely different things implying the existence of a domain-agnostic process. We then show that this holds for relatively difficult but not easy explanations.

Effect of Missing Marketing Promotions on Future Consumption

Ilana Shanks, Florida State University, USA

This research examines how missing a marketing promotion acts as a cue to consumers to increase future consumption. Two studies demonstrate that missing a marketing promotion results in consumers intending to increase their food consumption during the next visit. This effect is moderated by deprivation. Advances in Consumer Research (Volume 45) / 1077 When Small Predicts Large: The Effect of Initial Small Contributions on Subsequent Contributions to a Crowdfunding Project

Yael Steinhart, Tel Aviv University, Israel Leilei Gao, Chinese University of Hong Kong, China Tingting Fan, Chinese University of Hong Kong, China

Large donations are considered to be vital for crowdfunding success. Herein large-scale field data and lab experiments show that counterintuitively small (rather than large) initial donations increase subsequent donations. This effect is mediated by the lay belief that small (vs. large) donations are likely to be from strangers (vs. friends).

Effort Type Predicts Preferences for Material or Experiential Goods

Christopher A. Summers, University of South Carolina, USA Eva C. Buechel, University of South Carolina, USA Gustavo Schneider, University of South Carolina, USA

We propose that effort type can predict consumers’ preferences for rewards. In three studies we show that physical effort leads to a greater preference for material rewards whereas mental effort leads to a greater preference for experiential rewards. Our results identify effort type as an antecedent to consumer preference.

“Broken Eggs” and “Good Eggs”: Understanding “Post-IVF” Consumer Ambivalence Through Autography

Jennifer Takhar, Novancia Business School, France

This visual and narratological analysis of IVF consumption through autography/visual memoir demonstrates the multiple representational advantages afforded by the genre for consumer research; notably it allows privileged access to elusive experiential consumer “veraci- ty”. It also elucidates deep consumer ambivalence about IVF that is indissociable from its normalization as a technology.

The Merits of Happy Consumption: Positive Affect and Psychological Ownership

Carina Thürridl, Wirtschafts University, Austria Bernadette Kamleitner, Wirtschafts University, Austria Ruta Ruzeviciute, Wirtschafts University, Austria Stephan Dickert, Queen Mary University of London, UK Sophie Süssenbach, Wirtschafts University, Austria

The feelings consumers experience during consumption can have powerful effects but can they also influence how possessive one feels towards the consumed? In this paper we examine whether positive affect experienced during brand consumption can instill a sense of ownership for the consumed brand subsequently leading to intended repeat consumption.

Preemptive Social Influence: (Not) Choosing Personal Favorites in Shared Consumption?

Yijie Wang, Hong Kong Polytechnic University, Hong Kong Dongjin He, Hong Kong Polytechnic University, Hong Kong Yuwei Jiang, Hong Kong Polytechnic University, Hong Kong

Consumers are less likely to choose their personal favorite option when making decisions for shared consumption. This effect is weak- ened when consumers believe that they are similar to others or there is a high-power distance in the society and is strengthened when they share consumption with their close friends. 1078 / Working Papers Author Index

Barauskaitė, Dovilė ...... 1014 Blocker, Christopher ...... 707 A Barbosa, Belem ...... 785 Blümle, Sarah ...... 1058 Aaker, Jennifer ...... 63, 79, 129 Bardhi, Fleura ...... 1049, 1058 Bock, Dora ...... 858 Abell, Annika ...... 906, 1073 Barger, Victor A...... 47 Bodenhausen, Galen ...... 595 Abella, Mireya Sosa ...... 914 Barnes, Aaron J...... 515 Bodur, H. Onur ...... 1025, 1041 Abi, Gulen Sarial ...... 1055 Barnett, Samuel B...... 383 Bolderdijk, Jan Willem ...... 1016 Abi, Gülen Sarial ...... 865, 867 Barnhart, Michelle ...... 675 Bollinger, Bryan ...... 291 Abrams, Deborah ...... 1011 Barone, Michael J...... 324, 851 Bond, Sam ...... 1050 Acar-Burkay, Sinem ...... 1011 Barrios, Andres ...... 1008 Bond, Samuel ...... 1011 Achar, Chethana ...... 36, 493, 665 Bartels, Daniel M...... 96, 346 Bone, Sterling ...... 1008 Acikalin, M. Yavuz ...... 42 Barth, Madeline ...... 518 Bonnemaizon, Audrey ...... 379 Agrawal, Nidhi ...... 36, 493, 665, 890 Barth-Strieder, Kerstin ...... 520 Booyse, Wim ...... 1015 Ahluwalia, Rohini ...... 850 Baskentli, Sara ...... 206, 233 Borah, Abhishek ...... 140 Akdevelioglu, Duygu ...... 1011 Baskin, Ernest ...... 47, 112, 839 Borenstein, Benjamin ...... 1015 Alemany-Oliver, Mathieu ...... 496 Bastos, Wilson ...... 1014 Borges, Adilson ...... 970, 1029 Almeida, Stefânia Ordovás de . . . . . 1002 Basu, Shankha ...... 525 Bos, Maarten ...... 276 Alós-Ferrer, Carlos ...... 264 Batat, Wided ...... 1014, 1014 Botelho, Delane ...... 1024 Alter, Adam ...... 1055 Batra, Rajeev ...... 1046 Botner, Keith ...... 42 Alter, Adam L...... 270 Batra, Rishtee ...... 618 Boztug, Yasemin ...... 683 Amatulli, Cesare ...... 498 Baumeister, Roy ...... 1040 Brace-Govan, Jan ...... 1027 Amendah, Eklou ...... 1011 Baxter, Stacey ...... 527 Bradford, Tonya Williams ...... 57 Amine, Abdelmajid ...... 379 Beal, Alec ...... 654 Brakus, J Joško ...... 539 Amir, On ...... 196, 196, 212, 238, 373 Bebek, Gaye ...... 387 Brandes, Leif ...... 1007, 1016 Amri, Nuha Nasser Salem Al ...... 1001 Bech-Larsen, Tino ...... 1014 Brannon, Daniel ...... 540, 1016 An, Donghwy ...... 1012, 1012 Beck, Joshua ...... 217 Brei, Vinicius ...... 1051 An, Jiyoon ...... 1012 Beck, Joshua T...... 74 Brick, Danielle J...... 362 Anderson, Laurel ...... 1008 Bee, Colleen ...... 1047 Briley, Donnel ...... 1070 André, Quentin ...... 170, 336 Bei, Lien-Ti ...... 1015 Brito, Eliane ...... 1017 Angelis, Matteo De ...... 498, 578 Bekkers, Rene ...... 79 Brizi, Ambra ...... 539 Appau, Samuelson ...... 1008 Belei, Nina ...... 626, 629 Brokalaki, Zafeirenia ...... 1008 Aral, Sinan ...... 181 Belk, Russ ...... 1005 Brouwer, Claire ...... 1016 Argo, Jennifer . . . . .175, 1034, 1054, 1062 Belk, Russell . . . . 117, 164, 618, 677, 1005 Brown, Stephen ...... 397 Arias, Robert ...... 500 Bellezza, Silvia ...... 107, 558, 1005 Brucks, Melanie S...... 25 Ariely, Dan ...... 222, 731 Belli, Alex ...... 1073 Brucks, Merrie ...... 258 Aron-Wisnewsky, Judith ...... 572 Bellis, Emanuel de ...... 31 Bruin, Wändi Bruine de ...... 367 Arsenijevic, Vladan ...... 1013 Bennett, Delancy ...... 860 Bsdurek, Anja ...... 896 Arthur, Damien ...... 1001 Bentley, Kara ...... 530 Buechel, Eva ...... 888 Aschemann-Witzel, Jessica ...... 1054 Berendt, Johannes ...... 924 Buechel, Eva C...... 1077 Ashraf, Rohail ...... 1019 Berger, Jonah . . . . .52, 107, 248, 291, 1005 Bullard, Olya ...... 1054 Ashworth, Laurence ...... 829 Bergers, Verena ...... 1045 Bulmer, Sandy ...... 588 Atakan, S. Sinem ...... 503 Bergh, Bram Van den ...... 1031 Burgh-Woodman, Helene de ...... 401 Atalay, Selin ...... 1018 Berman, Jonathan Z...... 281, 1006 Burson, Katherine ...... 645 Attari, Amin ...... 507, 1013 Berman, Ron ...... 52 Burton, Scot ...... 1021 Audrezet, Alice ...... 509 Berry, Christopher ...... 1008 Büttner, Oliver ...... 286 Auschaitrakul, Sumitra . 1013, 1039, 1039, 1050 Bettman, James ...... 233 Austin, Caroline Graham ...... 1013 Bettman, James R. . . . .264, 362, 1005, 1006 Ay, Ufuk ...... 1062 Beuckels, Emma ...... 532 C Ayaz, Hasan ...... 1015 Bezençon, Valéry ...... 787 Cabano, Frank ...... 507 Aybat, Ozge ...... 91 Bharadwaj, Anupama Mukund . . . . . 1050 Cacabelos, Rosa ...... 1016 Aydınoğlu, Nilüfer ...... 1018 Bharadwaj, Sundar ...... 158 Cai, Fengyan ...... 1067, 1073 Bhatia, Sudeep ...... 346 Calderon, Susan ...... 545 Bhatt, Siddharth ...... 1015 Caliandro, Alessandro ...... 1016 B Bhattacharjee, Amit ...... 534 Camanho, Nelson ...... 547 Bagchi, Rajesh ...... 243, 932, 1067 Bhattacharya, CB ...... 206 Cambefort, Marine ...... 549 Bagozzi, Richard P...... 1067, 1067 Bhogal-Nair, Anoop ...... 394 Cambier, Fanny ...... 551 Bai, Yuxin ...... 511 Bi, Sheng ...... 801 Camilleri, Adrian ...... 1016 Ballantine, Paul W...... 1038 Biraglia, Alessandro ...... 539 Camilleri, Adrian R...... 553, 1056 Banai, Ádám ...... 809 Biswas, Dipayan . . .233, 313, 313, 906, 1073 Campana, Mario ...... 1016 Banerjee, Akshina ...... 512 Bittner, Samantha ...... 1015 Campbell, Colin ...... 1030, 1042 Banker, Sachin ...... 514 Blank, Ashley Stadler ...... 900 Campbell, Margaret C...... 1 Banovic, Marija ...... 1013 Bleier, Alexander ...... 140 Campbell, Troy ...... 842, 1047 Barakshina, Tatiana ...... 36 Block, Lauren ...... 91, 233, 1006, 1008 Canli, Zeynep Gürhan ...... 867 Barasch, Alixandra . . . .68, 123, 1056, 1059 Blocker, Chris ...... 1008 Advances in Consumer Research 1079 Volume 45, ©2017 1080 / Author Index

Cao, C. Clark ...... 258 Cheng, Zhao-Hong ...... 732 Dalli, Daniele ...... 406 Cao, Clark ...... 1007 Chin, Ching ...... 1021 Dalton, Amy N...... 228 Cao, Fei ...... 1017 Chiu, Tseng-Ping ...... 1020 Daniels, Michelle ...... 1023 Cardoso, Flavia ...... 1017 Cho, Eunsun ...... 1057 Dannewald, Till ...... 683 Carlson, Kurt ...... 679, 756 Cho, Yoon-Na ...... 1021, 1021 Davidson, Alexander ...... 1023 Carlson, Kurt A...... 852 Choe, Yuna ...... 1021 Davis, Brennan ...... 1051 Carmon, Ziv ...... 212 Chou, Hsuan-Yi ...... 1021 Davis, Scott W...... 629 Carnevale, Marina ...... 313 Chowdhry, Nivriti ...... 42 Decrop, Alain ...... 1002 Carpenter, Gregory S...... 152, 578, 605 Christenson, Brett ...... 1021 Dellaert, Benedict G. C...... 158 Carrillat, François ...... 1073 Christopoulos, George ...... 1062 Dellaert, Benedict G.C...... 101 Carter, Erin Percival ...... 63 Chrysochou, Polymeros ...... 602 Demmers, Joris ...... 582 Carufel, Courtney A...... 68 Chu, Wujin ...... 1043 DeMotta, Yoshiko ...... 1024 Carvalho, Lilian ...... 1017 Chu, Xing-Yu ...... 563, 565, 1021 Deng, Qian (Claire) . . . . .1024, 1024, 1073 Castaño, Raquel ...... 200, 1007, 1046 Chuang, Yaliang ...... 1001 Deng, Xiaoyan . .175, 1006, 1035, 1036, 1039, Castelo, Noah ...... 1017, 1065 Chung, Hoeun ...... 1022 1049, 1060 Castilhos, Rodrigo ...... 1008 Chung, Jaeyeon ...... 107 Deng, Xun ...... 1024 Castilhos, Rodrigo B...... 135 Chung, Jung-Han ...... 1021 DePaoli, Alexander ...... 583, 1024 Catapano, Rhia ...... 63 Chung, Myungjin ...... 571 Desai, Preyas ...... 36 Cauberghe, Verolien ...... 532 Cian, Luca ...... 286, 1006 Deutsch, Jonathan ...... 1015 Cauberghe, Veroline ...... 1034 Clair, Julian Saint ...... 860 DeVoe, Sanford ...... 63 Caucigh, Elodie ...... 1007 Clarkson, Joshua ...... 795, 1051 Dewhirst, Timothy ...... 1025 Cavanaugh, Lisa A...... 31, 1006, 1075 Clarkson, Joshua J...... 74 Deyl, Fernanda Trindade ...... 1062 Cecutti, Lorenzo ...... 1017 Clément, Karine ...... 572 Dhal, Darren ...... 1009 Cerf, Moran ...... 383, 1007 Clement, Michel ...... 270 Dhar, Ravi ...... 25, 96, 158, 264 Cesareo, Ludovica ...... 555, 558 Cléret, Baptiste ...... 1005 Dhar, Rhavi ...... 212 Ceylan, Melis ...... 1018 Clergue, Valentina ...... 1007 Dholakia, Utpal ...... 42 Chabot, Aimee ...... 79 Cocker, Hayley ...... 511 Dickert, Stephan ...... 1077 Chae, Boyoun (Grace) ...... 561 Cole, Cathy ...... 1036 Diehl, Kristin ...... 31, 341 Chae, Rebecca ...... 1018 Colliander, Jonas ...... 1003 Dietvorst, Berkeley J...... 302 Chai, Luxi ...... 1018, 1018 Cook, Laurel ...... 1002 Dimofte, Claudiu ...... 1043 Chakravarti, Amitav ...... 999 Corciolani, Matteo ...... 52 Dimoka, Angelika ...... 1007 Chan, Cindy ...... 270, 302, 1046 Cordeiro, Rafaela Almeida ...... 1011 Ding, Isabel ...... 1035, 1036, 1074 Chan, Eugene ...... 948 Cordes, Henning ...... 1022 Ding, Ying ...... 243, 1029 Chandler, Vincent ...... 1051 Corneille, Olivier ...... 1066 Ding, Yu ...... 585, 990 Chandon, Pierre ...... 572, 574 Cornelis, Erlinde ...... 769 Distler, Elisabeth ...... 1048, 1049 Chang, Chia-Chi ...... 1018 Cornelissen, Gert ...... 1016 Djedidi, Amina ...... 406 Chang, Chun-Tuan ...... 563, 565, 732 Cornil, Yann ...... 572, 574 Dobele, Angela R...... 1056 Chao, Matthew ...... 346 Cornwell, T. Bettina ...... 1022 Dodds, Sarah ...... 588 Chaplin, Lan Nguyen ...... 1009 Cossu, Martina ...... 1022 Doering, Tim ...... 1006 Chartrand, Tanya ...... 961 Costa, Sandrine ...... 1025 Dolbec, Pierre-Yann ...... 590 Chartrand, Tanya L...... 362, 974 Costabile, Michele ...... 578 Dolen, Willemijn M. van ...... 582 Chatterjee, Promothesh ...... 507, 1013 Costello, John ...... 1023 Dommer, Sara ...... 458 Chatterjee, Subimal . . . . .1018, 1057, 1059 Cotte, June ...... 951, 1049 Dommer, Sara Loughran ...... 200 Chatterjee, Swagato ...... 1019 Coulter, Robin ...... 1007 Dong, Ping ...... 307 Cheatham, Lauren ...... 152, 1019 Coulter, Robing A...... 57 Donnelly, Grant E...... 191, 319 Chen, Annie ...... 1019, 1053 Cowart, Kelly ...... 1053 Dorri, Sadaf Mokarram ...... 865 Chen, Fangyuan ...... 1033, 1068 Craig, Adam ...... 1007 Dose, David B...... 874 Chen, Huan ...... 801, 1073 Craig, Adam W...... 297 Douglas, Brownlie ...... 1005 Chen, Jia ...... 1020 Crain, Katherine M...... 264 Dovgialo, Karolina ...... 1025 Chen, Jinjie ...... 68 Cristian, Daniela ...... 1023, 1023 Drake, Carly ...... 410 Chen, Lin-Lin ...... 1001 Critcher, Clayton ...... 217 Drakopoulos, Stavros ...... 1041 Chen, Ning (Chris) ...... 1019 Crosby, Elizabeth ...... 1008 Drozdova, Natalia ...... 415 Chen, Po-Yu ...... 1018 Cross, Samantha ...... 1026 Drumwright, Minette ...... 854 Chen, Qihui ...... 1020 Cryder, Cynthia ...... 68 Du, Rui ...... 1025 Chen, Qimei ...... 1025 Cryder, Cynthia E...... 898, 1006 Du, Shuili ...... 206 Chen, Rocky Peng ...... 350, 1068 Curasi, Carolyn ...... 955 Duan, Meng ...... 694 Chen, Rong ...... 964 Cutright, Keisha ...... 146, 555, 561 Dubois, David ...... 578, 1007, 1041 Chen, Rui ...... 567, 1019 Czellar, Sandor ...... 1041, 1041 Duffy, Katherine ...... 1016 Chen, Shijiao ...... 569, 1020 Duhachek, Adam ...... 324, 727 Chen, Shirley Shuo ...... 1020 Duke, Kristen ...... 196 Chen, Yen-Ting ...... 732 D Dunfield, Derek ...... 514 Chen, Yu-Jen ...... 1020 Dagogo-Jack, Sokiente ...... 217 Dunn, Elizabeth ...... 63 Chen, Yung-Hsun ...... 1001 Dahl, Darren ...... 970 Dunn, Elizabeth W...... 1006 Chen, Yu-Shan Athena ...... 1001, 1015 Dahl, Darren W...... 212, 362 Dunn, Lea ...... 36, 1050 Chen, Zengxiang ...... 1045, 1075 Dai, Hengchen ...... 302 Durante, Kristina ...... 1009 Cheng, Li-Keng ...... 1015 Dai, Xianchi ...... 723, 885, 1059 Duval, Kimberly ...... 1025 Cheng, Shirley Y Y ...... 725 Dallas, Steven ...... 222, 576 Dyen, Margot ...... 1025 Cheng, Yimin ...... 36 Dallas, Steven K...... 91 Advances in Consumer Research (Volume 45) / 1081

Gaeth, Gary J ...... 479 Gray, Kurt ...... 228, 1065 E Gal, David ...... 341, 764, 1005 Grayson, Kent ...... 52, 152, 607, 764 Ebrahimi, Mahdi ...... 79 Galak, Jeff ...... 112, 217, 276 Greer, Dominique ...... 1055 Eckles, Dean ...... 181 Gallage, H.P. Samanthika ...... 419 Gretry, Anaïs ...... 626, 629 Ein-Gar, Danit ...... 253, 1009 Galli, Maria ...... 1033 Grewal, Dhruv ...... 47, 112 Ekpo, Akon ...... 1026 Galoni, Chelsea ...... 107, 248, 605, 607 Grewal, Lauren ...... 112, 140 Ellen, Pam ...... 955 Gao, Fei ...... 1028 Griesser, Simone ...... 1030 Emontspool, Julie ...... 438 Gao, Hongzhi ...... 569, 1020 Grigsby, Jamie L...... 1030 Eng, Teck Y...... 1007 Gao, Huachao ...... 1027 Grillo, Tito L. H...... 1030 Erdem, Tulin ...... 286 Gao, Leilei ...... 101, 908, 1077 Grimes, Jerry ...... 1030 Erner, Carsten ...... 1022 Garbarino, Ellen ...... 609 Grimm, Pamela ...... 1042 Esch, Patrick van ...... 928 Garbinsky, Emily N...... 612 Grishin, Marija ...... 634, 1018 Escoe, Brianna ...... 962 Garcia-Rada, Ximena ...... 123, 222 Griskevicius, Vladas ...... 85, 253 Espartel, Lélis Balestrin ...... 1062 Gardner, Meryl ...... 1009, 1067 Groeppel-Klein, Andrea . . . . 636, 640, 896 Estes, Zachary ...... 286, 313, 1022 Gauri, Dinesh ...... 170 Grohmann, Bianca ...... 620 Etkin, Jordan . . . . . 79, 291, 291, 346, 946 Geiger-Oneto, Stephanie ...... 614 Grønhøj, Alice ...... 1031 Evangelidis, Ioannis ...... 593 Geisler, Ally ...... 928 Grossman, Daniel ...... 641 Evers, Ellen R.K...... 276 Germelmann, Claas Christian ...... 1060 Grunert, Klaus G...... 1014 Gerrath, Maximilian ...... 1028 Gu, Yangjie ...... 101 Gershoff, Andrew ...... 152, 1043 Guèvremont, Amélie ...... 642, 1031 F Gershoff, Andrew D...... 778 Guido, Gianluigi ...... 498 Fajardo, Tatiana ...... 258, 270, 1015 Gershon, Rachel ...... 68 Gullo, Kelley ...... 146, 291, 362 Falkowski, Andrzej ...... 1035 Geyskens, Kelly ...... 1058 Gunadi, Manissa ...... 1031 Fan, Linying (Sophie) ...... 85, 330 Ghaddar, Suad ...... 469 Gunadi, Manissa P...... 644 Fan, Tingting ...... 1077 Ghosh, Anastasiya Pocheptsova . 74, 803, 1034 Gunasti, Kunter ...... 799 Fan, Xiaomeng ...... 595, 1073 Ghoshal, Tanuka ...... 618 Guo, Zhaoyang ...... 1071 Fan, Yafeng ...... 1026, 1026, 1029 Giblin, Colleen ...... 1028 Gupta, Reetika ...... 1031 Farah, Maya ...... 596 Gier, Nadine ...... 1028, 1041 Gurdamar, Nazli ...... 281 Farmer, Matthew ...... 598, 1026 Giesler, Markus ...... 1007 Gürhan-Canli, Zeynep . . . . 703, 1032, 1037 Fedorikhin, Alexander ...... 1035 Gilad, Zohar ...... 1074 Feick, Lawrence ...... 743 Gill, Dr. Tripat ...... 1052 Fennis, Bob ...... 856, 1023, 1023, 1028 Gineikiene, Justina ...... 600, 1028 H Fennis, Bob M...... 600, 1011, 1014 Gineikienė, Justina ...... 1014 Ha, Young-Won ...... 1021, 1045 Ferguson, Melissa ...... 91 Gino, Francesca ...... 191, 720 Haasova, Simona ...... 1031 Ferguson, Shelagh ...... 1027 Gino, Frencesca ...... 716 Habibi, Mohammad Reza ...... 1023 Fernances, Daniel ...... 789 Giroux, Marilyn ...... 620, 892 Hadi, Rhonda ...... 313 Fernandes, Daniel ...... 547, 780 Givi, Julian ...... 112, 217 Hagen, Linda ...... 645 Ferreira, Maura ...... 1074 Gladstone, Joe J...... 612 Hagtvedt, Henrik ...... 258 Festila, Alexandra ...... 602 Gloukhovtsev, Alexei ...... 1028 Hagvtedt, Henrik ...... 1006 Fischer, Eileen ...... 1007 Gneezy, and Ayelet ...... 238 Hahn, Minhi ...... 1038 Fishbach, Ayelet 91, 129, 146, 181, 248, 356, 714 Gneezy, Ayelet ...... 191 Hajjat, Fatima ...... 1031 Fisher, Eileen ...... 57 Go, Sarang ...... 1058 Hajjat, Fatima M...... 648 Fisher, Robert J...... 729 Godfrey, D. Matthew ...... 117, 135 Hall, Matthew J...... 650 Fitzsimons, Gavan . . 146, 222, 222, 233, 1056 Goedegebure, Robert P.G...... 622 Hall-Phillips, Adrienne ...... 1011 Fitzsimons, Gavan J. . .91, 107, 362, 974, 1006 Goldenberg, Jacob ...... 1056 Halvari, Halgeir ...... 1037 Flaschen, Katherine ...... 1074 Goldsmith, Emily ...... 1029, 1030 Hamilton, Mitchell ...... 860 Fleck, João Pedro ...... 1002 Goldsmith, Kelly . . .85, 248, 253, 1029, 1057 Hamilton, Rebecca ...... 1007, 1009 Florack, Arnd ...... 1031 Goldstein, Daniel G...... 336 Hamilton, Rebecca W...... 140, 362 Flores, Miriam C...... 1064 Golf-Papez, Maja ...... 624 Hamilton, Ryan ...... 101, 158, 270, 822 Folkes, Valerie ...... 1063 Gomez, Pierrick ...... 1029 Hampton, William ...... 367 Folse, Judith Anne Garretson ...... 264 Gonçalves, Dilney ...... 545 Han, Bing ...... 1032, 1032 Forcum, Lura ...... 603 Gong, Han ...... 669 Han, DaHee ...... 324 Ford, John ...... 826 Gong, Xiushuang ...... 1029 Han, Jerry ...... 152, 652 Franck, Egon ...... 1016 Gonzales, Gabriel ...... 1029 Han, John J...... 281 Frederick, Shane ...... 96 Goode, Miranda ...... 951 Handelman, Jay M...... 135 Freichel, Markus ...... 640 Goodman, Joseph K...... 123, 228 Hanson, Sara ...... 1012 Friedman, Liz ...... 96, 158 Goodyear, Laura ...... 1029, 1057 Harden, Drew ...... 1025 Frizzo, Francielle ...... 1027 Gopinath, Mahesh ...... 1061 Hardisty, David ...... 63, 654 Frow, Pennie ...... 1070 Goranson, Amelia ...... 228 Hassan, Louise ...... 1032 Fu, Pei-Wen ...... 1027 Gordeliy, Ivan ...... 1042 Hattula, Johannes ...... 25 Fujita, Kentaro ...... 1060 Gorn, Gerald ...... 1032 Häubl, Gerald ...... 101, 158, 158, 186 Fulton, Benjamin ...... 1015 Gorn, Gerald J. . . . .1034, 1059, 1060, 1060 Hauser, Oliver ...... 191 Fumagalli, Elena ...... 1074 Goswami, Indranil ...... 96, 129 Haws, Kelly ...... 146, 233, 1008 Goukens, Caroline ...... 1058 Hayran, Ceren ...... 1032 Gould, Stephen ...... 1029, 1030 He, Daniel ...... 1033 G Gram, Malene ...... 1031 He, Dongjin ...... 1032, 1033, 1077 Grant, Annetta ...... 135 Gaertig, Celia ...... 1027 He, Jibo ...... 1046 Graul, Antje ...... 1063 Gaeth, Gary ...... 1036, 1059 He, Sharlene ...... 152 1082 / Author Index

He, Xin ...... 753 Huh, Kyoung Tae ...... 1035 Johnson, Eric J...... 367 Heath, Teresa ...... 419 Huh, Young Eun ...... 1034 Johnson, Russell E...... 663 Hedgcock, Bill ...... 1007 Humayun, Mariam ...... 117, 677 Joseph, Joshy ...... 465 Heger, Stephanie ...... 609 Humphrey, Colman ...... 52 Josion-Portail, Margaret ...... 379 Heideker, Silvia ...... 655 Humphreys, Ashlee ...... 52 Jr., John G. Lynch ...... 170 Hellén, Katarina ...... 657 Hurwitz, Jonathan ...... 1033 Jr., Marcus Cunha ...... 567 Heller, Jonas ...... 928 Hutchinson, J. Wesley ...... 297 Jun, Sunghee ...... 1037 Hem, Leif ...... 766 Hwang, Heungsun ...... 449 Jun, Youjung ...... 181 Hemetsberger, Andrea ...... 474, 1040 Hydock, Chris ...... 679 Jung, Minah H...... 270 Henry, Paul ...... 1005 Hyodo, Jamie D...... 650 Jung, Sung J...... 1075 Herd, Kelly ...... 1006 Hyun, Yong J...... 1052 Jungfleisch, Hannah ...... 896 Herd, Kelly B...... 31 Juntunen, Jouni ...... 481 Herpen, Erica van ...... 622 Herrmann, Andreas ...... 31, 158 I Hershfield, Hal E...... 79, 248 Ibrahim, Nahid ...... 158 K Herter, Márcia Maurer ...... 1022, 1048 Iglesias, Claudia ...... 1035 Kahn, Barbara E...... 291 Herzenstein, Michal ...... 52 II, Americus Reed ...... 350 Kähr, Andrea ...... 1037 Herzog, Walter ...... 25 Ilicic, Jasmina ...... 527 Kaju, Alex ...... 698, 768 Heuvinck, Nico ...... 769 Imas, Alex ...... 238 Kamleitner, Bernadette . . . . 313, 699, 1077 Hieke, Silke ...... 31 Inbar, Yoel ...... 307 Kan, Christina ...... 1021 Hietanen, Joel ...... 1003, 1005 Ince, Elise Chandon ...... 803, 932 Kang, Christine ...... 1018, 1044 Hildebrand, Christian . . . 31, 158, 164, 1074 Inman, J. Jeffrey ...... 291, 720 Kang, Dawon ...... 1043 Hiler, Jacob ...... 1002 Inman, Jeff ...... 297 Kang, Jungyun ...... 1038 Hill, Sarah ...... 1009 Irmak, Caglar ...... 930 Kang, Seo Yoon ...... 1037 Hills, Thomas T...... 1030 Islek, Mahmut Sami ...... 1062 Kangaspunta, Klaus ...... 1003 Hingston, Sean T...... 659 Ito, Kenichi ...... 31 Kapitan, Sommer ...... 1055 Hmurovic, Jillian ...... 291, 661 Ittersum, Koert van ...... 856 Kapoor, Ankur ...... 701, 916 Ho, Cony M...... 663 Ittersum, Koert Van ...... 901 Kara, Selcan ...... 1011, 1064 Hoang, Chi ...... 1033 Iversen, Nina ...... 1035 Karantinou, Kalipso ...... 785 Hock, Stefan J...... 243 Iversen, Nina M...... 766 Karatas, Mustafa ...... 703, 940, 1037 Hoegg, JoAndrea ...... 561, 970, 1055 Iyer, Easwar ...... 989 Kardas, Michael ...... 123 Hoegg, JoAndrea (Joey) ...... 1006 Izadi, Anoosha ...... 175 Kardes, Frank ...... 976 Hoelzl, Erik ...... 264 Kardes, Frank R...... 962 Hoffman, Donna ...... 164 Karmarkar, Uma R...... 96 Hoffman, Donna L...... 117 J Kasani, Negin Latifi ...... 1043 Hofstetter, Reto ...... 341 Jablonska, Magdalena ...... 1035 Kazakova, Snezhanka ...... 532 Holmes, Cassie Mogilner . . . 79, 1006, 1007 Kedzior, Richard ...... 1038 Homonoff, Tatiana ...... 170 Jahn, Steffen ...... 683 Jain, Gaurav ...... 479, 1059 Keenan, Elizabeth A...... 191, 238 Hong, Sujeong ...... 1033 Keinan, Anat ...... 1005 Hong, Ying-Yi ...... 1065 Jain, Shailendra ...... 976 Jain, Shailendra P...... 1047 Kelleher, Carol ...... 705 Hong, Yoonah ...... 1034 Keller, L. Robin ...... 948 Horvath, Csilla ...... 626, 629 Jakubanecs, Alexander ...... 1035 Jami, Ata ...... 685, 716 Keller, Sarah ...... 1013 Hossain, Shajuti ...... 553 Kelley, Corinne ...... 206 Hosseini, Rahil ...... 1033 Janakiraman, Narayan ...... 977 Janiszewski, Chris ...... 356 Kelly, Brooke ...... 307 Hoyer, Wayne ...... 140 Kemper, Joya A...... 1038 Hoyer, Wayne D...... 1037 Jarry, Francis ...... 1064 Javornik, Ana ...... 423 Kenning, Peter . . . . 1028, 1041, 1045, 1058 Hsee, Christopher ...... 228 Kerrane, Ben ...... 433 Hsee, Christopher K...... 68, 196 Jayarajah, Kasthuri ...... 1031 Jenkins, Adrianna ...... 667 Kerviler, Gwarlann Caffier de ...... 509 Hsieh, Meng-Hua ...... 493, 665 Khan, Fatima Yaqub ...... 428 Hsu, Ming ...... 196, 667 Jeon, Eunmi ...... 1038 Jeong, Haeyoung ...... 350 Khan, Jashim ...... 1053 Hsu, Shu-Ni ...... 1069 Khan, Sadia Yaqub ...... 428 Hsu, Wei-Chin ...... 1015 Jewell, Robert D...... 1030, 1048 Jia, Jayson S...... 42 Khan, Uzma ...... 42, 583 Hu, Miao ...... 1025 Kharuhayothin, Tanyatip ...... 433 Hu, Yu ...... 1033 Jia, Lei ...... 175, 1006, 1035, 1036 Jia, Miaolei (Liam) . 233, 286, 964, 1035, 1036 Kidwell, Blair ...... 707 Hu, Yuheng ...... 341 Kidwell, Virginie Lopez ...... 707 Huang, Alex ...... 514 Jia, Yanli ...... 687 Jiang, Jing ...... 1026, 1026, 1066, 1067 Kim, Byungdo ...... 1058 Huang, Feifei ...... 671, 1066 Kim, Claire Heeryung ...... 324, 727 Huang, Li ...... 673, 1034, 1034, 1034 Jiang, Yuwei . . 85, 330, 330, 671, 1006, 1032, 1033, 1059, 1060, 1060, 1077 Kim, Eunjin (Anna) ...... 1039 Huang, Liang ...... 74 Kim, Hakkyun ...... 1038 Huang, Szu-chi . . . . . 91, 129, 1049, 1074 Jiao, Jinfeng (Jenny) . . 330, 1036, 1036, 1046 Jie, Joseph Yun ...... 689 Kim, Janghyun ...... 1012 Huang, Yanliu ...... 291 Kim, Jin M...... 1075 Huang, Yunhui ...... 669, 692 Jin, Daoyan ...... 1037 Jin, Fei ...... 1036 Kim, Joonkyung ...... 1039, 1053 Huang, Zhongqiang (Tak) ...... 1034 Kim, Joshua ...... 248 Huber, Frank ...... 520 Jin, Liyin ...... 692 Jin, Zhenyu ...... 694 Kim, Junhee ...... 350 Hudders, Liselot ...... 532, 1034 Kim, Kaeun ...... 1039 Huettel, Scott ...... 1056 Johar, Gita V...... 107 John, Deborah Roedder ...... 68, 1009 Kim, Kyeongheui ...... 1044 Huff, Aimee ...... 675 Kim, Min Jung ...... 1038 Hughes, Mine Üçok ...... 486 John, Leslie ...... 68 John, Leslie K...... 319, 341 Kim, Nicole Y...... 74, 362 Advances in Consumer Research (Volume 45) / 1083

Kim, Pielah ...... 1039 Levontin, Liat ...... 253, 1009, 1074 Kim, Sang-Hoon ...... 1035 L Levy, Eric ...... 350 Kim, Sara ...... 350, 350, 1038 Labroo, Aparna ...... 777 Li, Charis ...... 238 Kim, Seung Eun (Sonia) ...... 1075 Labroo, Aparna A...... 107 Li, Eric ...... 1005, 1044, 1044 Kim, Yaeri ...... 1053 Ladeira, Wagner Junior ...... 580 Li, Fiona ...... 1044 Kim, Youngsoo ...... 1038 Lafreniere, Katherine C...... 729 Li, Keyi ...... 1036 King, Dan ...... 1013, 1039, 1039 Lai, Ai-Ling ...... 443 Li, Qi ...... 1044 Kirk, Colleen P...... 712 Lalwani, Ashok ...... 734, 736 Li, Shaobo ...... 206, 1075 Kirmani, Amna ...... 1005 Lam, Howard Pong-Yuen ...... 1066 Li, Xilin ...... 228 Kitayama, Shinobu ...... 1020 Lam, Magnum ...... 1044 Li, Xingbo ...... 324, 1075 Kivetz, Ran ...... 1033 Lamberton, Cait ...... 661, 718, 872 Li, Xiuping ...... 206, 233, 286 Kladou, Stella ...... 1053 Lambotte, Soraya ...... 1038 Li, Ye ...... 367, 742 Kleintop, Lizabeth ...... 928 Lamprinakos, Grigorio ...... 785 Li, Yexin Jessica ...... 634, 985, 1018 Kleppe, Ingeborg A...... 415 Landgraf, Polina ...... 1042 Li, Yuan ...... 740 Kliebenstein, Stephanie ...... 640 Landy, Justin F...... 212 Liang, Jianping ...... 1045 Knight, John ...... 569, 1020 Lane, Kristen ...... 200 Liang, Yitian (Sky) ...... 1034 Knoferle, Klemens ...... 1033 Langer, Thomas ...... 1022 Lieberman, Alicea J...... 25, 212 Knutson, Brian ...... 367 Langhe, Bart de ...... 336 Lim, Elison ...... 1075 Kobel, Sarah ...... 636 Langmaack, Ann-Christin ...... 270 Lim, Mikyoung ...... 1045 Kocher, Bruno ...... 1007, 1041, 1041 Lanseng, Even ...... 1042 Lim, Ming ...... 443 Koenigstorfer, Joerg ...... 1047 Laran, Juliano . .85, 175, 356, 356, 1009, 1015, Lim, Sarah ...... 228 Koetz, Clara ...... 1058 1040 Lin, Chia-Wei Joy ...... 743, 1045 Kogler, Stephanie ...... 1040 Larkin, Ian ...... 346 Lin, Chien-Wei (Wilson) ...... 747, 749 Kokkinaki, Flora ...... 1041, 1055 Laroche, Michel ...... 1023, 1070 Lin, Meng-Hsien (Jenny) ...... 1064 Kokkoris, Michail D...... 206, 264, 1040 Larrick, Richard P...... 553 Lin, Ruoyun ...... 217, 1027, 1045 Konya-Baumbach, Elisa ...... 518 Laurent, Gilles ...... 1041 Lin, Shan ...... 1045 Koo, Minjung ...... 714, 1044 Leban, Marina ...... 1043 Lin, Stephanie ...... 307 Koo, Minkyung ...... 1073 LeBoeuf, Robyn A...... 898 Lin, Szu-Han (Joanna) ...... 663 Kooi, Kate ...... 1040 Lee, Angela ...... 1009 Lindridge, Andrew. M...... 394 Kordrostami, Elika ...... 826, 1040 Lee, Angela Y ...... 253 Lindsey, Charles ...... 976 Kordrostami, Melika ...... 1064 Lee, Angela Y...... 1047 Ling, Christopher ...... 751 Korelo, Jose Carlos ...... 1027 Lee, Bo Han ...... 1057 Lischka, Helena M...... 1045 Kosinski, Michal ...... 276 Lee, Boram ...... 1012 Lisjak, Monika ...... 1064 Kouchaki, Maryam ...... 685, 716 Lee, Chang-Yuan ...... 731 Litovsky, Yana ...... 281 Koulaei, Afra ...... 1040 Lee, Chulsung ...... 1012 Liu, Fan ...... 753 Kousi, Sofia ...... 1041 Lee, Hsiao-Ching ...... 732 Liu, Jingshi (Joyce) ...... 228 Kovacheva, Aleksandra ...... 718 Lee, Hsin-Hsuan Meg ...... 454 Liu, Kailuo ...... 1046 Kozinets, Robert ...... 1005 Lee, Hyejin ...... 734, 736 Liu, Maggie Wenjing ...... 1071 Kramer, Thomas ...... 751 Lee, Hyojin ...... 1060 Liu, Martin ...... 1007 Krampe, Caspar ...... 1028, 1041 Lee, Jacob ...... 1044 Liu, Peggy . . 47, 112, 146, 146, 222, 233, 576 Krastel, Zachary ...... 1041 Lee, Jaewoo ...... 1044 Liu, Wendy ...... 196, 373, 997 Kraus, Karolina ...... 438 Lee, Jeff ...... 1007, 1042 Liu, Wing-sun ...... 1044 Krekels, Goedele ...... 1041, 1041 Lee, Jennifer K...... 31, 1075 Liu, Wumei ...... 1046 Kreps, Tamar A...... 307 Lee, Jeonggyu ...... 1015 Liu, Xiaoyan ...... 1075 Krishna, Aradhna . . 233, 286, 286, 755, 1006 Lee, Jeongmin ...... 1043 Liu, Xuefeng ...... 1046 Kristofferson, Kirk . . .212, 1009, 1023, 1063 Lee, Jin-Myong ...... 1057 Liu-Thompkins, Yuping ...... 1040 Krohmer, Harley ...... 1037 Lee, Jiyoung ...... 1043 Loewenstein, George ...... 341, 876 Kronrod, Ann ...... 1042 Lee, Jonathan ...... 449 Loken, Barbara ...... 900, 968 Kruger, Liezl-Marié ...... 57 Lee, Kyoungmi ...... 1075 Longoni, Chiara ...... 47, 286 Ksendzova, Masha ...... 186, 191 Lee, Leonard . . . . . 206, 1008, 1068, 1074 Lopez, Alberto ...... 1046 Kuchmaner, Christina ...... 1042, 1042 Lee, Margaret ...... 742 Loughlin, Deirdre M. O’ ...... 705 Kühnen, Ulrich ...... 1040 Lee, Seojin stacey ...... 1053 Loveland, Katherine E...... 755 Kumar, Anand ...... 297 Lee, Seojin Stacey ...... 1053 Lowe, Michael ...... 458, 755, 1011 Kumar, Ankita ...... 135 Lee, Seung Yun ...... 778 Lowengart, Oded ...... 238 Kumar, Anvita ...... 1026 Lee, Shinhyoung ...... 1043 Lowrey, Tina M...... 1028 Kumar, Ashish ...... 481 Lee, Wonkyong Beth ...... 1025 Lteif, Lama ...... 91 Kupor, Daniella ...... 42, 181, 196, 835 Lee, Yih Hwai ...... 129, 1068 Lu, Fang-Chi ...... 1036, 1046 Kurt, Didem ...... 720, 959 Lee, Youseok ...... 1035, 1058 Lu, Joy ...... 96 Kurultay, Ayse Binay ...... 537 Lefebvre, Sarah ...... 313 Lu, Wei ...... 694 Kwak, Hyokjin ...... 350, 995 Lei, Jing ...... 1045 Lu, Zoe Y...... 756 Kwan, Canice M C ...... 725 Leischnig, Alexander ...... 1048, 1049 Luangrath, Andrea Webb ...... 47 Kwan, Canice M. C...... 723 Lemaire, Alain ...... 52 Lubowski, Ruben ...... 654 Kwon, Eunseon (Penny) ...... 1039 Lembregts, Christophe ...... 644 Luca, Ramona De ...... 1024 Kwon, JaeHwan ...... 727 Leong, Josiah K...... 367 Luce, Mary Frances ...... 36, 264 Kwong, Jessica ...... 1017 Leung, Ada ...... 738 Lucke, Sabrina ...... 1047 Leung, Eugina ...... 164 Luethi, Anja ...... 1049 Levav, Johnathan ...... 164 Luna, David ...... 313, 793 Levav, Jonathan ...... 25, 96 Lunardo, Renaud ...... 758, 1054 Levin, Irwin P ...... 479 1084 / Author Index

Luo, Jin ...... 694 Melumad, Shiri ...... 25, 52 Munz, Kurt P...... 270 Luo, Xiyueyao ...... 1075 Mende, Martin ...... 206, 1008 Murphy, Andrew ...... 588 Lupoli, Matt ...... 238 Mendenhall, Zachary ...... 1039 Murphy, Mary C...... 74 Luri, Ignacio ...... 57, 117, 762 Meng, Hua (Meg) ...... 1048 Murray, Jeff ...... 1008 Lusch, Robert F...... 117 Meng, Rachel ...... 181 Murray, Kyle B ...... 1020 Lutz, Richard ...... 1007 Meng, Yan ...... 1048 Lyle, Keith B...... 324 Meng, Zhu ...... 85 Lynch, John G...... 780 Menon, Geeta ...... 47 N Messer, Uwe ...... 1048, 1049 Naderi, Iman ...... 330, 1050 Messinger, Paul ...... 1024, 1024, 1073 Nagpal, Anish ...... 1060 M Mette, Frederike Monika Budiner . . . . .580 Nardini, Gia ...... 373 Ma, Dr. Zhenfeng ...... 1052 Meyer, Robert ...... 52 Nave, Gideon ...... 276 Ma, Jingjing 764, 1020, 1020, 1047, 1047, 1075 Meyers-Levy, Joan ...... 968 Nayakankuppam, Dhananjay . .479, 727, 1059 Ma, Yoon Jin ...... 1052 Meyvis, Tom ...... 186 Nelson, Leif ...... 1076 Ma, Yu ...... 170 Miceli, Gaetano ...... 771 Nelson, Leif D...... 276 Ma, Zhenfeng ...... 1047 Mick, David ...... 1007, 1008 Netemeyer, Richard ...... 780 MacInnis, Debbie ...... 21 Miller, Chadwick ...... 458 Netzer, Oded ...... 52, 96, 302 Macinnis, Deborah ...... 1007 Miller, Elizabeth ...... 1031, 1039 Neville, Fergus G...... 1060 Maclaran, Pauline ...... 397 Miller, Elizabeth G ...... 989 Newman, and George E...... 238 Madrigal, Robert ...... 842, 1047 Miller, Elizabeth G...... 648 Newman, George E...... 264 Madzharov, Adriana ...... 1006 Mimoun, Laetitia ...... 773, 1049 Ng, Sharon ...... 525 Maehle, Natalia ...... 766, 1037 Min, Bora ...... 152 Nicolaides, Christos ...... 181 Maeng, Ahreum ...... 243 Min, Dong-Jun ...... 1049 Nicolao, Leonardo . . . . . 1022, 1048, 1051 Maglio, Sam ...... 248 Min, Hyun Jeong ...... 775 Nielsen, Jesper ...... 598, 872, 1069 Maglio, Sam J...... 698, 768 Min, Kate ...... 1044 Nielsen, Jesper H...... 200 Mai, Li-Wei ...... 1053 Min, Kyeong Sam ...... 1049 Niemiec, Christopher ...... 1037 Maimaran, Michal ...... 74 Mingolla, Carla ...... 1034 Nikolova, Hristina ...... 140, 458 Main, Kelley ...... 1054 Minton, Elizabeth ...... 614 Nolte, Sven ...... 1051 Main, Kelley J...... 1054 Mishra, Arul ...... 42, 831 Nordgren, Loran F...... 848 Majmundar, Anuja ...... 769 Mishra, Himanshu ...... 42, 831 Northey, Gavin ...... 928 Malkoc, Selin A...... 123 Mittal, Chiraag ...... 85, 253, 1009 Northington, William ...... 1002 Malter, Alan ...... 36 Mittal, Vikas ...... 894 Norton, Michael ...... 63, 79 Mandel, Naomi ...... 540, 1064, 1067 Mo, Tingting ...... 1049 Norton, Michael I. . . . . 123, 186, 228, 1006 Mannetti, Lucia ...... 539 Mo, Zichuan ...... 1075 Noseworthy, Theodore ...... 1006 Mao, Huifang ...... 1036 Mochon, Daniel ...... 196 Noseworthy, Theodore J...... 659 Margolin, Drew ...... 1070 Moerk, Trine ...... 1014 Novak, Thomas ...... 164 Mari, Silvia ...... 1067, 1067 Mogilner, Cassie ...... 63, 302 Novak, Thomas P...... 117 Marozzo, Veronica ...... 771 Mohammad, Aylar Pour ...... 1055 Nowlan, Luke ...... 175, 1015, 1051 Marques, Thiago Rafael Ferreira . . . . 1001 Mohr, Gina S...... 952 Nowlan, William ...... 1040 Martin, Brett A.S...... 699 Moldovan, Sarit ...... 884 Nowlis, Stephen M...... 898 Martin, Diane ...... 1027 Monga, Ashwani ...... 888, 1072 Nuñez, Sandra ...... 200 Martin, Kelly D...... 57 Montaguti, Elisa ...... 1063 Nyffenegger, Bettina ...... 1037 Martin, Olga ...... 1047 Mookherjee, Satadruta ...... 1018 Martinovici, Ana ...... 286 Moon, Alice ...... 276 Marzooqi, Sara Mohamed Al ...... 1001 Moore, Sarah G...... 222 O Moorhouse, Michael ...... 1049 Mas, Erick ...... 707 O’Brien, Ed ...... 123, 782 Mason, Marlys ...... 824 Morales, Andrea ...... 1009, 1023 Morales, Andrea C...... 212, 362 O’Brien, Rourke ...... 170 Masset, Julie ...... 1002 O’Donnell, Michael ...... 276 Mastrogiannis, Dimitrious ...... 36 Moreau, Page ...... 31, 31, 1060 Morewedge, Carey ...... 1028 O’Loughlin, Deirdre M...... 785 Mather, Damien ...... 569, 1020 Oh, Hyewon ...... 1051 Mathras, Daniele ...... 1008 Morewedge, Carey K...... 186, 222, 731 Morgan, Carter ...... 1051 Olivola, Christopher ...... 191, 217 Mathur, Pragya ...... 91 Olivola, Christopher Y...... 281, 281 Matz, Sandra ...... 276 Morgan, Samuel Carter ...... 258 Morgan, Siân ...... 74 Olson, Jenny G...... 362, 634, 1006 Mazodier, Marc ...... 91 Omeira, Mansour ...... 787 McAlexander, Brandon ...... 675 Mormann, Milica ...... 297 Morrin, Maureen ...... 233, 1006, 1069 Ong, L. Lin ...... 1008 McAlexander, Jim ...... 675 Ordabayeva, Nailya ...... 140, 789, 1007 McColl-Kennedy, Janet ...... 1008 Morris, Joshua ...... 1049 Morvinski, Coby ...... 238, 373 Ordenes, Francisco Villarroel ...... 47 McEachern, Morven G...... 785 Orsingher, Chiara ...... 791 McFerran, Brent . . . . .146, 212, 362, 1009 Moulard, Julie Guidry ...... 264, 509 Mourali, Mehdi ...... 1061, 1070 Ostberg, Jacob ...... 1005 McGill, Ann ...... 362 Ostinelli, Massimiliano ...... 1069 McGowan, Miriam ...... 1032 Mourey, James A...... 777 Moya, Maria Eugenia Fernández . . . . .785 Ostinelli, Max ...... 793 McShane, Lindsay ...... 1076 Otnes, Cele ...... 500, 766, 1051 Mead, Nicole L...... 330 Mukherjee, Ashesh . . . . . 778, 1013, 1050 Mukherjee, Sourjo ...... 1031 Otterbring, Tobias ...... 1076 Mehta, Ravi ...... 175, 1006, 1051 Otto, Ashley ...... 795, 1051 Mellers, Barbara ...... 346 Mukherjee, Sudipta ...... 1050, 1050 Mukhopadhyay, Anirban ...... 36, 228 Ourahmoune, Nacima ...... 406 Mello, Rafael ...... 1002 Ozanne, Lucie ...... 1008 Meloy, Margaret ...... 1029 Mulier, Lana ...... 1050 Muller, Brigitte ...... 1041 Ozcan, Timucin ...... 799 Meloy, Margaret G...... 852 Ozmen, Mujdat ...... 1062 Advances in Consumer Research (Volume 45) / 1085

Poole, Maxwell ...... 1051, 1076 Rocklage, Matthew D...... 848 P Popovich, Deidre ...... 822 Rodas, Maria ...... 850 Packard, Grant ...... 52, 1035 Pounders, Kathrynn ...... 824 Rodas, Maria A...... 200, 851 Paik, Sung-Hee ...... 1022 Powell, Ashleigh E...... 1056 Rodriguez-Vila, Omar ...... 158 Paik, Wendy ...... 903 Powell, Emily ...... 1055, 1056 Roeder, Scott ...... 1076 Paley, Anna ...... 85 Prado, Paulo Henrique ...... 1027 Rojas-Gaviria, Pilar ...... 1017 Palihawadana, Dayananda ...... 926 Prakash, Mona ...... 1056 Rokka, Joonas ...... 1003, 1040 Pancer, Ethan ...... 1051, 1076 Prasad, Ajnesh ...... 833 Romero, Marisabel ...... 297 Panchal, Shirish ...... 1052 Prelec, Drazen ...... 514 Romero-Canyas, Rainer ...... 654 Pandelaere, Mario . . . . . 1006, 1050, 1050 Price, Linda ...... 1007, 1008 Rosa, Jose A...... 1048 Pang, Jun ...... 801, 878, 1073 Price, Linda L...... 117 Rosario, Ana Babic ...... 1013 Paolacci, Gabriele ...... 164, 534 Prochaska, Judith J...... 807 Rosenboim, Mosi ...... 123 Park, Hye Kyung ...... 714 Prokopec, Sonja ...... 1061 Rosenzweig, Emily ...... 112, 217 Park, Jihye ...... 1052, 1053 Pundak, Chen ...... 1056 Ross, Gretchen ...... 852 Park, Jooyoung ...... 1052, 1052 Puntoni, Stefano ...... 164 Roth, Yefim ...... 1057 Park, Kiwan ...... 1037, 1053, 1053 Putnam-Farr, Eleanor ...... 212 Rothschild, David ...... 336 Park, Sungjun (Steven) ...... 1052 Puzakova, Marina ...... 350, 995 Roux, Caroline . . .85, 253, 1029, 1057, 1066 Park, Sunny ...... 1057 Pyone, Jin Seok ...... 1069 Roy, Rajat ...... 1057 Park, Taehoon ...... 803 Rozin, Paul ...... 307 Park, Yookyung ...... 1052 Rucker, Derek D...... 324, 324, 848 Parker, Andrew M...... 367 Q Rudd, Melanie ...... 79, 175 Parker, Jeffrey R...... 158, 200 Qin, Vivian ...... 107 Ruiz, Veronica Martin ...... 1048 Patino-Echeverri, Dalia ...... 553 Ruppell, Roland ...... 341 Patrick, Vanessa ...... 1062, 1070 Ruttan, Rachel ...... 818 Patrick, Vanessa M...... 79, 175, 258, 993 R Ruvio, Ayalla ...... 1005 Paul, Iman ...... 158, 200 Ruzeviciute, Ruta ...... 313, 1077 Paulson, Erika L...... 805 Radford, Scott ...... 410 Ryoo, Yuhosua ...... 854 Pearson, John ...... 1056 Rafieian, Hoori ...... 291 Pechmann, Cornelia ...... 1008 Raggio, Randle D...... 264 Pechmann, Cornelia (Connie) ...... 807 Raghubir, Priya ...... 297 S Peck, Joann ...... 47, 712, 818 Raghunathan, Raj ...... 1050 Saad, Gad ...... 1017 Pecot, Fabien ...... 549 Rahinel, Ryan ...... 641, 795 Sääksjärvi, Maria ...... 657 Peleg, Gil ...... 238 Rahmani, Vahid ...... 826, 1040 Saccardo, Silvia ...... 191, 191, 373 Pelet, Jean-Eric ...... 1053 Rai, Dipankar ...... 747 Sadowski, Sebastian ...... 856 Pellandini-Simányi, Léna ...... 809 Raimondo, Maria Antonietta ...... 771 Saenger, Christina ...... 858 Pelsmacker, Patrick De ...... 532 Rajagopal, Priyali ...... 530, 673 Sagfossen, Sofie ...... 1003 Peng, Norman ...... 1019, 1053 Ramadan, Zahy ...... 596 Sah, Sunita ...... 319 Penner, Sara ...... 1054, 1054 Rank-Christman, Tracy ...... 1076 Sahay, Arvind ...... 701 Peracchio, Laura ...... 1006, 1008 Rao, Hayagreeva ...... 605 Saini, Ritesh ...... 571, 977 Pereira, Beatriz ...... 812 Ratchford, Mark ...... 827 Saintives, Camille ...... 758 Perfecto, Hannah ...... 276 Rath, Suzanne ...... 829 Salerno, Anthony . . . . .253, 270, 356, 1009 Perkins, Andrew ...... 801 Rathee, Shelly ...... 831 Salo, Jari ...... 481 Perren, Rebeca ...... 1054 Ratner, Rebecca ...... 123, 1009 Samanez-Larkin, Greg ...... 367 Peschel, Anne Odile ...... 1054 Ratner, Rebecca K...... 362 Sampaio, Claudio Hoffmann ...... 580 Petit, Olivia ...... 1054 Rauf, Ateeq Abdul ...... 833 Samper, Adriana . . . . .146, 152, 270, 1063 Petsonk, Annie ...... 654 Read, Daniel ...... 319 Sanglé-Ferrière, Marion ...... 862 Pew, Ethan ...... 281, 827 Rech, Eduardo ...... 1056 Santini, Fernando de Oliveira ...... 580 Pham, Michel T...... 25 Reczek, Rebecca Walker ...... 1023 Saridakis, Charalampos ...... 926 Phillips, Megan ...... 1055 Redden, Joseph ...... 233 Savani, Krishna ...... 206, 319, 585 Philp, Matthew ...... 829 Redden, Joseph P...... 186 Savary, Jennifer . . . . . 158, 200, 238, 1026 Phipps, Marcus ...... 1008 Reghunathan, Aravind ...... 465 Sayarh, Nada ...... 1057 Pierre-Yann, Dolbec ...... 815 Reich, Brandon ...... 842, 1047 Scaraboto, Daiane ...... 1017 Pieters, Rik ...... 286 Reich, Taly ...... 248, 307, 835, 839 Schau, Hope ...... 762 Pino, Giovanni ...... 498 Reichelt, Valesca Persch ...... 580 Schau, Hope J...... 57 Pinto, Diego Costa ...... 1022, 1048 Reid, Colbey Emerson ...... 1006 Schau, Hope Jensen ...... 117, 1007 Pizzetti, Marta ...... 423 Reimann, Martin . .200, 258, 1006, 1007, 1046, Schembri, Sharon ...... 469 Pizzutti, Cristiane . . .1030, 1056, 1058, 1074 1058 Scherer, Fernanda ...... 1058 Plank, Andreas ...... 461 Reinholtz, Nicholas ...... 96, 170, 336 Scherr, Karen ...... 233 Plassmann, Hilke ...... 572, 1007 Reynolds, Thomas ...... 844 Schley, Dan R...... 336 Poehlman, T. Andrew ...... 1065 Rha, Jong-Youn ...... 1057 Schlosser, Ann ...... 140 Pogacar, Ruth ...... 313 Rick, Scott ...... 812, 937, 1006 Schmidt, Liane ...... 572 Polman, Evan ...... 222, 768, 818 Rickard, Bradley J...... 1054 Schmitt, Bernd ...... 31, 1017 Polyakova, Alexandra ...... 791 Rifkin, Jacqueline ...... 291 Schneider, Gustavo ...... 1077 Polyportis, Athanasios ...... 1055 Ringler, Christine ...... 1021 Schneider, Judith ...... 1051 Ponchio, Mateus ...... 1011 Risen, Jane L...... 319 Schoenmueller, Verena ...... 302 Poncin, Ingrid ...... 551 Ritter, Ryan ...... 228 Scholz, Joachim ...... 869 Ponsonby-McCabe, Sharon ...... 397 Robertson, Kirsten ...... 1061, 1061 Schöps, Jonathan D...... 474 Pontes, Nicolas ...... 1055, 1055, 1076 Robitaille, Nicole ...... 829 Schouten, John ...... 1003, 1005 Pontes, Vivian ...... 1055, 1076 Roche, Kévin ...... 1065, 1066 1086 / Author Index

Schreier, Martin ...... 31 Smith, Andrew ...... 869 Takhar, Jennifer ...... 1077 Schreiner, Nadine ...... 1058 Smith, Cristalle ...... 1044 Talebi, Arash ...... 1061 Schrift, Rom Y...... 373 Smith, Robert ...... 186, 782 Talloen, Joachim ...... 196 Schroeder, Juliana ...... 25 Smith, Rosanna ...... 835 Tan, Teck Ming ...... 481 Schultz, Ainslie ...... 872 Smith, Rosanna K...... 264, 839 Tang, Chuanyi ...... 1061 Schumacher, Anika ...... 1058 So, Jane ...... 890 Tanner, Emily C...... 1008 Schwabe, Maria ...... 874 Soares, André Escórcio ...... 1026 Tanner, Robin ...... 756 Schwartz, Daniel ...... 238, 876 Sobande, Francesca ...... 773 Tanner, Robin J...... 243 Schwartz, Janet ...... 36 Sobol, Kamila ...... 892, 1057 Tao, Tao ...... 101, 908 Schwarz, Norbert ...... 152, 966 Söderlund, Magnus ...... 1003 Teodorescu, Kinneret ...... 1057 Scopelliti, Irene ...... 341 Soltwisch, Brandon ...... 1016 Terres, Mellina ...... 1022, 1048 Scott, Maura ...... 1006, 1006, 1008 Som, Anirban ...... 1059 Tezer, Ali ...... 85, 1029 Scott, Maura L...... 206 Song, Eunyoung ...... 1044 Theotokis, Aristeidis ...... 926 Scott, Sydney ...... 307 Song, Jiaqi ...... 1059, 1060, 1060 Thomas, Manoj ...... 170 Scott, Sydney E...... 212 Song, Peijian ...... 1044, 1071 Thomas, Tabitha ...... 1061, 1061 Seaman, Kendra ...... 367 Song, Xiaobing ...... 894 Thomas, Veronica L...... 858 Sekhon, Tejvir ...... 1012 Soster, Robin ...... 1021 Thompson, Craig J...... 6, 135 Sela, Aner ...... 74, 248, 373 Souissi, Fayrouz ...... 1052 Thompson, Debora ...... 1009 Sen, Sankar ...... 206 Soule, Catherine Armstrong . . . .1012, 1012 Thürridl, Carina ...... 699, 1077 Seok, Junhee ...... 1058 Spiller, Stephen A...... 1006 Thyne, Maree ...... 1061, 1061 Seregina, Anastasia ...... 1005 Spilski, Anja ...... 896 Tignor, Stefanie M...... 910 Serial-Abi, Gülen ...... 101 Srinivasan, Raji ...... 140 Tillotson, Jack ...... 1062 Setten, Eric ...... 1022 Srna, Shalena ...... 68, 146 Tiltay, Muhammet Ali ...... 1062 Sevilla, Julio ...... 253, 1006, 1009 Staelin, Richard ...... 107 Ting, Ding Hooi ...... 914 Shah, Anuj ...... 1009 Stahl, Florian ...... 302, 1063 To, Ngoc (Rita) ...... 258, 1062 Shah, Avni M...... 1006 Stamatogiannakis, Antonios . . 545, 980, 1042 To, Rita ...... 1006 Shanks, Ilana ...... 1076 Stavros, Constantino ...... 1056 Tok, Dickson ...... 563, 565 Shankweiler, James ...... 738 Steffel, Mary ...... 112, 319 Tonietto, Gabriela ...... 123 Shapira, Danny ...... 238 Steinhart, Yael ...... 884, 1056, 1077 Torelli, Carlos J...... 200, 324, 851 Sharifonnasabi, Zahra ...... 1058 Steinmetz, Janina ...... 248 Torlak, Omer ...... 1062 Sharma, Eesha ...... 85, 1009 Stephen, Andrew T...... 140 Tormala, Zakary ...... 152, 181, 1019 Sharma, Varun ...... 313 Stephenson, Brittney L...... 898 Tormala, Zakary L...... 42 Shavitt, Sharon ...... 324 Steul-Fischer, Martina ...... 655 Torres, Lez Trujillo ...... 773 Shehu, Edlira ...... 270 Stevens, Lorna ...... 397 Toure-Tillery, Maferima ...... 248 Sheinin, Daniel ...... 1012 Stewart, Kristin ...... 1054 Townsend, Claudia ...... 258, 1006 Shen, Hao ...... 687, 878, 1019 Stieler, Maximilian ...... 1060 Tran, An ...... 449 Shen, Luxi ...... 196, 336, 879, 1075 Stillman, Paul E...... 1060 Trapido, Denis ...... 807 Sheng, Feng ...... 196 Stillwell, David ...... 276 Trendel, Olivier ...... 91, 1065 Sherry, John F...... 57 Stone, Aaron ...... 1068 Trijp, Hans C.M. van ...... 622 Sheth, Jagdish ...... 17 Stoner, Jennifer L...... 324, 900 Tripathi, Sanjeev ...... 916 Shibly, Sirajul ...... 1059 Stough, Rusty ...... 1060 Trudel, Remi ...... 191, 910, 959 Shim, Yoonji ...... 881 Stovall, Tony ...... 486 Truong, Natalie ...... 1034, 1062 Shiu, Edward ...... 1032 Streicher, Mathias ...... 286 Tsai, Claire ...... 652, 918 Shiv, Baba ...... 42, 196 Strejcek, Brendan ...... 607 Tsalis, George ...... 1014 Shoham, Meyrav ...... 884 Strelow, Enrique ...... 1041 Tsang, Alex S L ...... 725 Shrivastava, Sunaina ...... 479, 1059 Strough, JoNell ...... 367 Tu, Lingjiang Lora ...... 922 Shrum, L.J...... 1074 Stuhler, Kristina ...... 1061 Tu, Ping ...... 1036 Shukla, Paurav ...... 1007 Su, Lei ...... 330, 1066 Tu, Yanping ...... 146, 181, 228 Si, Kao ...... 885, 1059 Suher, Jacob ...... 901 Tully, Stephanie M...... 85 Siddiqui, Rafay ...... 74, 1072 Sullivan, Nicolette ...... 1056 Ture, Meltem ...... 503 Siddiqui, Rafay A...... 888 Summers, Christopher A...... 1077 Tynan, Caroline ...... 419 Silverman, Jackie ...... 1059 Sun, Daniel ...... 1061 Simmons, Joseph ...... 1027 Sun, Sijie ...... 977 Simmons, Joseph P...... 276 Sundar, Aparna ...... 903 U Simonson, Itamar ...... 302, 373, 583 Suri, Rajneesh ...... 1015 Ubel, Peter ...... 576 Sinai, Olivier ...... 1014 Süssenbach, Sophie ...... 1077 Uhalde, Arianna ...... 1063 Singh, Surendra ...... 985, 1013, 1069 Sussman, Abigail ...... 170 Uhrich, Sebastian ...... 924 Singh, Surendra N...... 634 Sussman, Abigail B...... 346, 1006 Ulusoy, Emre ...... 1063 Sirianni, Nancy ...... 1021 Svirsky, Daniel A...... 319 University, Gita V. Johar Columbia . . . .181 Sirieix, Lucie ...... 1025 Swain, Scott D...... 712 Unnava, Rao ...... 1060 Skarsten, Sondre ...... 336 Swanepoel, Samantha ...... 1015 Upadhyaya, Shikha ...... 1008 Slabbinck, Hendrik ...... 1050 Szmigin, Isabelle T...... 785 Urminsky, Oleg 96, 101, 129, 297, 336, 512, 982 Slonim, Robert ...... 609 Szocs, Courtney . . . . .233, 901, 906, 1073 Usrey, Bryan ...... 926, 1028, 1063 Slot, Johanna ...... 1029 Szugalski, Stefan H...... 1003 Usta, Murat ...... 158 Sluis, Helen van der ...... 1063 Utochkin, Denis ...... 415 Small, Deborah ...... 68, 307, 1059 Utz, Sonja ...... 217, 1027 Smeets, Paul ...... 79 T Smidts, Ale ...... 1007 T, Aruna Divya ...... 1019 Advances in Consumer Research (Volume 45) / 1087

Warren, Caleb ...... 952, 1005 V Warren, Danielle ...... 1076 Y Valentini, Sara ...... 791 Warren, Nathan ...... 1047 Yamim, Amanda ...... 1065 Valenzuela, Ana ...... 1033, 1048 Waytz, Adam ...... 228 Yamim, Amanda Pruski ...... 970 Vallen, Beth ...... 1008 Weathers, Danny ...... 1065 Yan, Dengfeng ...... 740 Valli, Veronica ...... 1063 Weaver, Lisa ...... 738 Yan, Jun ...... 1068 Valsesia, Francesca ...... 341 Weaver, Todd ...... 955 Yang, Adelle ...... 982 VanBergen, Noah ...... 930 Webb, Elizabeth C...... 302 Yang, Adelle X...... 68, 362 Veer, Ekant ...... 624 Wegerer, Philipp ...... 489 Yang, Chun-Ming . . . 972, 1019, 1068, 1069 Velez, Caty ...... 1064 Wegerer, Philipp K...... 474 Yang, Haiyang ...... 980, 1042 Velloso, Luciana ...... 57 Wei, Chuang ...... 1071 Yang, Li ...... 1068 Veludo-de-Oliveira, Tania ...... 1001 Wei, Sarah ...... 101, 186 Yang, Linyun W...... 974 Ven, Niels Van De ...... 217 Weinberg, Charles B...... 1034 Yang, Tzu-Yun ...... 1069 Venkatraman, Vinod ...... 96, 367, 987 Weingarten, Evan ...... 297, 346 Yang, Xiaojing 175, 694, 976, 1006, 1035, 1036, Ventzislavov, Rossen ...... 486 Weiss, Liad ...... 957 1069 Vermeir, Iris ...... 1050 Welles, Christoph ...... 1031 Yang, Yang ...... 181 Victor, Bart ...... 827 Werle, Carolina O.C...... 1065, 1066 Yang, Zheshuai ...... 129 Vijayalakshmi, Akshaya ...... 1064 Weseen, Sarah ...... 1044 Yang, Zhiyong ...... 977 Villanova, Daniel ...... 932 West, Colin ...... 63 Ye, Ning ...... 1069 Viswanathan, Madhu ...... 1008 Whillans, Ashley ...... 63, 79 Yeung, Catherine ...... 206 Vohs, Kathleen D...... 91, 258, 1006 White, Katherine ...... 881 Yi, Jisu ...... 1043 Vosgerau, Joachim ...... 341 White, Tiffany ...... 515 Yi, John ...... 1069 Voyer, Ben ...... 1025 Whitley, Sarah C...... 222, 959 Yi, Youjae ...... 1043, 1052 Voyer, Benjamin ...... 862, 1043 Wiener, Hillary ...... 961 Yim, Chi Kin (Bennett) ...... 243 Vredeveld, Anna ...... 1064 Wieser, Verena E...... 1066 Yin, Bingqing (Miranda) ...... 985 Vu, Tiffany ...... 937 Wiggins, Jennifer . . . . . 1042, 1042, 1065 Yin, Bingqing(Miranda) ...... 1069 Wijland, Roel ...... 397 Yoon, Carolyn ...... 1007, 1020 Wilcox, Keith ...... 356 Yoon, Heakyung Cecilia ...... 1037 W Williams, Elanor F...... 112, 319 Yoon, Heeyoung ...... 186 Wadhwa, Monica ...... 940, 943 Williams, Lawrence ...... 63 Yoon, Kelly EunJung ...... 807 Wakefield, Kirk ...... 297, 1051 Williams, Patti ...... 555 Yoon, Sangsuk ...... 987 Wallace, Scott ...... 346 Wilson, Anne ...... 253 Yorkston, Eric ...... 1056 Wallace, Scott G...... 946 Winkle, Jonathan ...... 1056 You, Yanfen ...... 1069 Wallendorf, Melanie ...... 135 Winterich, Karen ...... 1008 Youn, Nara . . . . . 1012, 1012, 1034, 1037 Waller, Daniel ...... 609 Winterich, Karen P...... 270, 324 Yu, Chao ...... 1070 Walsh, Darlene ...... 1064 Wolny, Julia ...... 1007 Yu, Sik Chuen ...... 1070 Walsh, Gianfranco ...... 874 Wong, Nancy ...... 1011 Yuksel, Mujde ...... 989 Walter, Maik ...... 158 Wong, Vincent Chi ...... 1066, 1066 Wan, Echo Wen ...... 243, 243, 330, 350 Woolley, Kaitlin ...... 129, 319, 356 Wan, Fang ...... 966, 1068, 1068 Wu, Charlene C...... 367 Z Wang, Carmen ...... 609 Wu, Chi-Cheng ...... 1027 Zamudio, Cesar ...... 1048 Wang, Haizhong ...... 1046 Wu, Eugenia ...... 222, 555, 718 Zamudio, César ...... 1065 Wang, Hsin-Yi ...... 565 Wu, Gavin Jiayun ...... 738 Zatz, Laura Y...... 319 Wang, Jing (Alice) ...... 330 Wu, Kaiyang ...... 243 Zaval, Lisa ...... 367 Wang, Juan ...... 951 Wu, Ruomeng ...... 962 Zavala, Mariella C...... 1070 Wang, Liangyan . . . .948, 1024, 1032, 1032 Wu, Yinghao ...... 1066, 1067 Zeng, Xianfang ...... 1070 Wang, Lili ...... 350 Wu, Yuechen ...... 101, 356, 362 Zerres, Alfred F...... 582 Wang, Lily ...... 1007 Wyer, Robert ...... 885 Zhang, Chun ...... 1070 Wang, Qi ...... 749 Wyer, Robert S...... 687, 723, 1065 Zhang, Jason ...... 1046 Wang, Qin ...... 948, 1064 Zhang, Kuangjie ...... 943 Wang, Qing ...... 1030 Zhang, Meng ...... 671 Wang, Quansheng ...... 1044, 1071 X Zhang, Weiwei ...... 1028 Wang, Ryan Tzushuo ...... 968 Xiao, Chunqu ...... 1064 Zhang, Yan ...... 129, 990 Wang, Shr-Chi ...... 563 Xie, Chunyan ...... 1007, 1067, 1067 Zhang, Ying ...... 91 Wang, Wangshuai ...... 1065 Xie, Jia lin ...... 918 Zhang, Yinlong . . .743, 894, 922, 1027, 1045 Wang, Xia ...... 1017 Xie, Jieru ...... 1067 Zhang, Yirang ...... 1071 Wang, Xian ...... 1070 Xie, Yi ...... 1067 Zhang, Yiwei ...... 170 Wang, Xin ...... 1064, 1071 Xiong, Ji (Jill) ...... 1068 Zhang, Yuan ...... 1071 Wang, Xue ...... 1065 Xu, Alison Jing ...... 324, 968 Zhang, Yuli ...... 995 Wang, Yajin ...... 68, 74, 1005 Xu, Huimin ...... 738 Zhang, Zhe ...... 993, 1006, 1070 Wang, Yijie ...... 1032, 1077 Xu, Jialiang ...... 966, 1068 Zhao, Xian ...... 1071 Wang, Yiru ...... 1065 Xu, Lidan ...... 175 Zhao, Xiaohua ...... 997 Wang, Ze ...... 753, 1017 Xu, Minzhe ...... 297 Zhao, Xin ...... 511, 1044, 1044 Ward, Adrian ...... 1034, 1065 Xu, Xiaobing ...... 964 Zhao, Yingnan ...... 1071 Ward, Adrian F...... 1030 Xue, Ke ...... 1032, 1032 Zheng, Frank ...... 1034 Ward, Morgan K...... 1005 Zheng, Xiaoying ...... 1071 Warlop, Luk ...... 1023, 1023, 1033 Zheng, Yuhuang ...... 966 Warmath, Dee ...... 780 Zhong, Chenbo ...... 1068 1088 / Author Index

Zhong, Chen-Bo ...... 307 Zhu, Huawei ...... 1036 Zimmermann, Laura ...... 999 Zhou, Lingrui ...... 146 Zhu, Luke ...... 1068 Zor, Ozum ...... 1072 Zhou, Xinyue ...... 350 Zhu, Meng ...... 243, 356 Zou, Lili Wenli ...... 243 Zhu, Hong ...... 1064 Zhu, Qichao ...... 1071 Zwebner, Yonat ...... 373 Association For Consumer Research

$99.00 ISBN 978-0-915552-77-1 59900>

9 780915 552771