Using Construal Level Theory to Deter the Social Desirability Bias

A dissertation submitted to the

Division of Research and Advanced Studies Of the University of Cincinnati

in partial fulfillment of the requirements for the degree of

DOCTORATE OF PHILISOPHY (PH.D.)

in the department of Marketing of the College of Business

2012

Scott A. Wright

B.S., Boston College

Committee Members: Frank R. Kardes, Ph.D., James J. Kellaris, Ph.D., David J. Curry,

Ph.D., and Rebecca Walker Naylor, Ph.D.

ABSTRACT

Leveraging Construal Level Theory to Attenuate the Social Desirability Bias

Scott A. Wright

Chair: Frank R. Kardes, Ph.D.

The social desirability bias is the tendency of respondents to adjust their responses or behavior in such a way as to present themselves in socially acceptable terms. The systematic bias introduced by the social desirability bias threatens the legitimacy of empirical research by confounding a phenomenon of interest with impression management behavior, thus obscuring research results and potentially triggering false conclusions. When the social desirability bias is a concern (e.g., when asking socially sensitive, embarrassing, or private questions), researchers commonly use techniques such as indirect questioning to avoid this bias. By asking respondents how most or the typical person would respond (i.e, referring to a third party target) the respondent transcends from an egocentric focus on his or her own unflattering attitudes or behavior onto that of an ambiguous target. Given the target‘s indistinctness respondents project their own feelings, attitudes, behaviors, or beliefs when responding while remaining psychologically distanced from the true, yet socially undesirable response.

According to construal level theory this process of ―transcendence‖ is possible because individuals are able to form abstract mental construals (Liberman and Trope 1998; Trope and

Liberman 2003; Trope and Liberman 2010). This mental construal process is essential to recalling the past, empathizing with others, imagining what could have been, and visualizing future events. We propose that when presented with an indirect question, respondents increase their mental construal to project a response onto the third party target. As mental construals increase, individuals refocus from detailed, incidental features to central, fundamental

2 characteristics (Trope and Liberman 2010). Consequently, we propose that indirect questioning prompts respondents to deemphasize the contextual demand to engage in impression management behavior by increasing construal levels.

This research serves three primary purposes. First, we propose construal level theory as a theoretical explanation for indirection questioning. Second, we explore how psychological distances and construal level effectively reduce social desirability bias. Lastly, we test our theoretical model within a consumer choice context including sustainable products. This research has important methodological and substantive implications for marketers, researchers, and consumer psychologists. In addition, we integrate construal level theory with projective measurement theory, thus extending knowledge in both domains. This research provides researchers with a framework to explain and develop techniques that deter social desirability bias.

In Study 1, we investigate the association of seven commonly used marketing scales with measures of socially desirable responding. As a preliminary demonstration that increasing construal levels decreases the social desirability bias, we manipulate the psychological distance of these seven marketing scales to reduce their association with socially desirable responding measures. Results indicate that whereas indirect questioning increasing social or spatial distance reduces the association with desirable responding measures, indirect questioning increasing temporal distance magnifies this association. Study 2 further investigates the relationship between construal level and the social desirability bias through a product choice task involving sustainable and non-sustainable product option. Results indicate that increasing construal levels in an ostensibly unrelated task subsequently increases the likelihood that consumers will select the non-sustainable product. In Study 3, we extend these findings by exploring the mechanism

3 explaining the relationship between psychological distance and impression management behavior. We find that when demand biases are present, increasing the psychological distance of the choice task reduces impression management, consequently increasing the share of choice for the functionally superior, yet environmentally inferior product option.

∙

Scott A. Wright

DEDICATION

To my loving wife, Christine.

ACKNOWLEDGEMENTS

I would like to acknowledge those that have helped me complete this work. Firstly, I would like to thank my committee chair, Frank Kardes, for his wealth of knowledge and support.

Frank has always treated me with kindness and the utmost respect. I am privileged and honored to have such a terrific mentor. I would also like to thank my extraordinary doctoral committee.

I would like to thank Dave Curry for his keen intellect and advice. The department is fortunate to have such an exceptional and compassionate Ph.D. coordinator. Thank you to James

Kellaris for his honesty and patience. Thank you to Rebecca Naylor for inspiring this dissertation and for guiding its development. I would also like to thank our department head, Karen Machleit for her guidance throughout my time in the program. I would also like to thank the remaining faculty at the University of Cincinnati for your support and for the opportunity to complete this work. Lastly, I would like to thank my wife, Christine, for whom I have dedicated this dissertation, for her eternal support.

TABLE OF CONTENTS

CHAPTER ONE Introduction…………………………………………..……………1

CHAPTER TWO Literature Review…………………………………….....…………5

CHAPTER THREE

Study 1………………………………….……………………………………………..…………17

Method……………………………………………………………………………..…….18

Results……………………………………………………………………………………22

Discussion……………………………………………………………………………..…24

CHAPTER FOUR

Study 2…………………………………………….…………………………..…………………26

Method………………………………………………………………………………..….28

Results……………………………………………………………………………………29

Discussion………………………………………………………………………………..30

CHAPTER FIVE

Study 3…………………………………………….……………………………………………..33

Method………………………………………………………………………………..….34

Results……………………………………………………………………………………37

Discussion……………………………………………………………………………..…39

CHAPTER SIX

General Discussion…………………………….………………………………………………...41

References…………………………………………………………………………………..……48

Tables and Figures………………………………………………………………………...……..59

Appendix A Manipulations and Measures for Study 1…………………………..……70 Appendix B Manipulations and Measures for Study 2……………………………..…90

Appendix C Manipulations and Measures for Study 3………………………………..98 Appendix D Individual Difference Measures…………………………………….…..105

TABLES AND FIGURES

TABLE 1 SUMMARY OF ARTICLES DOCUMENTING THE SOCIAL DESIRABILITY BIAS (JCR, JMR, JM, JAMS: 1998-2011) ……………………..………60

TABLE 2 DESCRIPTIVE STATISTICS (STUDY 1)……………..…………………….…63

TABLE 3 SUMMARY OF STANDIZED TOTAL, DIRECT, AND INDIRECT EFFECTS FOR EACH PRODUCT CATEGORY (STUDY 3)...…………….…64

FIGURE 1 STUDY 1: PLOTS OF WITHIN EXPERIMENTAL CONDITION STANDARDIZED REGRESSION COEFFICIENTS OF SDE AND IM FOR SEVEN MARKETING SCALES……………………………………….…65

FIGURE 2 STUDY 3: INDIRECT-ONLY MEDIATED MODERATION MODEL FOR LAUNDRY DETERGENTS ……..…………………………………….…68

FIGURE 3 STUDY 3: INDIRECT-ONLY MEDIATED MODERATION MODEL FOR HAND SANITIZERS…………..……………………………………….…69

CHAPTER ONE

Introduction

Whether completing a marketing survey, purchasing a product, or sharing product information with a friend, consumers frequently alter their responses or behavior to maintain positive social impressions (Baumeister 1982). This behavior, identified as the social desirability bias, has been well documented in the literature (Crowne and Marlowe 1960; Maccoby and

Maccoby 1954; Acquisti, John, and Loewenstein 2012; Steenkamp, de Jong, and Baumgartner

2010). A respondent‘s motive to engage in socially desirable responding increases as their accurate and truthful response deviates from a standard response that they perceive as being socially optimal. For example, questions pertaining to socially sensitive behaviors like calorie consumption, charitable giving, recycling, and television viewing are all likely to elicit distorted responses towards a socially acceptable standard or away from a socially unacceptable standard.

Similarly, scales measuring socially sensitive constructs such as materialism (Mick 1996), green orientation (Luchs et al. 2010), impulsive purchasing (Rook 1987), and corporate social responsibility (Burton and Hegarty 1999), are also likely to result in biased responses. To date, a significant portion of our knowledge of socially sensitive constructs, and their causal structures, comes from such self-report measures (Rindfleisch et al. 2008). Unfortunately, the social stigma associated with such topics, and consequent response adjustment behavior, increases systematic response biases, thus distorting parameter estimates and compromising the validity of research conclusions.

Prior research documents the threats posed by the social desirability bias (see King and

Bruner 2000; Mick 1996; Steenkamp et al. 2010). These threats range from inflated measurement error to outright erroneous results (see Fisher 1993; Ganster, Hennessey, and

Luthans 1983). This bias threatens measurement validity and diminishes the ability to predict important occurrences and phenomena. Prediction is the basis of all measurement, and an explanation for its prevalence and significance (see Dawes and Smith 1985). The social desirability bias could explain the inaccuracy of many predictive models relying on the validity of self-report data. For example, social desirability bias could explain the imprecision of many sales forecasting models (Lusk, McLaughlin, and Jaeger 2007; Morrison 1979). Moreover, socially desirable responding may contribute to the Bradley effect, or the observed discrepancy between opinion polls and election outcomes (Couzin 2008). In sum, by ignoring this bias, we disregard the potential threat of misleading and systematically biased empirical data, thus, weakening our ability to make accurate predictions and decisions.

In response, researchers have urged the field to increase efforts to detect and eliminate the social desirability bias during the design, measurement, or analysis stages of the research process. Similarly, research on question design has long emphasized the importance of accounting for important contextual demands, recommending question wording, along with other techniques, as a means of avoiding demand biases (Bradburn, Sudman, and Wansink 2004;

Schuman and Presser 1981; Sudman, Bradburn, and Schwarz 1996; Weaver and Schwarz 2008).

One such technique is indirect questioning. Classic research on projective techniques, demonstrates that asking how the 'average person‘ or ‗friend' would respond to a given question, versus how the respondent would respond, decreases socially desirable responding. Indirect questioning allows respondents to remain disassociated from potentially unflattering responses, while projecting their own opinions and behaviors onto the referent target (Calder and Burnkrant

1977; Grubb and Stern 1971; Haire 1950).

Fisher (1993) first demonstrated the effectiveness of indirect questioning within marketing. Across three studies, Fisher (1993) demonstrated that consumers distorted their responses to direct questions measuring socially sensitive constructs. However, when indirect questioning was used, the extent of socially desirable responding declined and responses reflected the respondents‘ true beliefs and evaluations. A number of techniques, in addition to indirect questioning, have been shown to deter or correct for the social desirability bias (i.e., item randomized response, bogus pipeline, etc.). Although effective under certain circumstances, these techniques have been criticized for being too costly, difficult to implement, demanding of respondents, or ethically controversial to achieve broad acceptance (see Tourangeau and Yan

2007).

Despite its advantages, indirect questioning remains an underutilized technique in marketing, in part, due to its dependence on social distance as the projective mechanism. In the current research, we extend the applicability of indirect questioning by testing additional projective dimensions, beyond that of social distance. For this purpose, we introduce construal level theory as a theoretical framework explaining indirect questioning and apply this framework to propose additional projective techniques.

Construal level theory explains how psychological distance influences mental representation, judgment, and choice (Trope and Liberman 2003; Trope and Liberman 2010;

Trope, Liberman, and Wakslak 2007). Psychological distance can vary in terms of time, space, social distance, probability, or any dimension that removes consumers from focusing on themselves in the here and now. As psychological distance on any of these dimensions increases, consumers are more likely to represent objects and events in high-level, abstract, and global terms that emphasize fundamental properties (e.g., superordinate categories, key attributes,

3 primary reasons for performing an activity). Conversely, as psychological distance on any of these dimensions decreases, consumers are more likely to represent objects and events in low- level, concrete, and local terms that emphasize peripheral issues (e.g., subordinate categories, peripheral attributes, secondary reasons for performing an activity). Thus, framing a question in the third (vs. first) person perspective increases its social distance, consequently emphasizing the construal of abstract features (i.e., traits) while deemphasizing situational factors (i.e., demand biases).

To this point, the projective measurement and construal level literatures have developed independently. The present dissertation seeks to integrate these lines of research to enhance theory in both domains. Three experiments have been designed to investigate psychological distance as a projective measurement technique. In this research, we demonstrate that other dimensions of psychological distance, in addition to social distance, successfully attenuate the social desirability bias. We also extend these techniques to consumer choice tasks. These findings contribute to our current conceptualization of indirect questioning and projective measure. It is our intention that researchers will acknowledge the versatility and ease of indirect questioning by implementing these techniques in their own work.

Summary of Chapters

Chapter 1 provides a brief overview of the dissertation‘s purpose, which is to examine psychological distance as a projective measurement technique. Chapter 2 reviews the literature integrating indirect measurement techniques with construal level theory as well as the literature on the social desirability bias. Chapter 3 presents Studies 1 through 3 along with results. Finally,

Chapter 4 provides a discussion of the results as well as a look forward to future studies.

CHAPTER TWO

Literature Review

Measures of The Social desirability Bias

The social desirability bias is the tendency of respondents to adjust their response in such a way as to present themselves in socially acceptable terms. This bias results in an adjustment either towards a socially acceptable standard (i.e., ―fake good‖) or away from a socially unacceptable standard (i.e., ―fake bad‖) (Paulhus and Reynolds 1995). Early conceptualizations of the social desirability bias considered this construct to be unidimensional, encompassing one‘s dispositional tendency to provide overly positive self-descriptions, (i.e., to engage in impression management behavior), (Paulhus 2002). However, the inability of scales measuring the social desirability bias to demonstrate convergent and predictive validity provoked the formation of various two-factor models (Borkenau and Ostendorf 1989).

Such models refer to the two dimensions of the bias as alpha and gamma bias, conscious and unconscious moralistic response tendencies, or impression management and self-deceptive enhancement (Paulhus 1991). For this investigation, we adopt the conventional nomenclature established by Paulhus and Bruce (1992) and refer to these factors as self-deception (SDE) and impression management (IM). According to Zerbe and Paulhus (1987), self-deception is ―the conscious tendency to see oneself in a favorable light,‖ and impression management is ―the conscious presentation of a false front, such as deliberately falsifying test responses to create favorable impressions‖ (p. 253). The authors argue that whereas self-deception is dispositional and relatively invariant to situational demands impression management is state dependent and thus prone to confound self-report data. Scales designed to detect the social desirability bias measure one or both of these factors. Among these scales are the Marlowe-Crowne scale

(Crowne and Marlowe 1960), the Balanced Inventory of Desirable Responding (BIRD) scale

(Paulhus 1991), the Edwards SD scale (Edwards 1957), the Wiggins SD scale (Wiggins 1959), the Other-Deception Questionnaire (Sackeim and Gur 1979), the Self- Deception Questionnaire

(Sackeim and Gur 1979), the Martin-Larsen Approval Motivation Scale (Larsen et al. 1976), and various EPI lie scales (see Steenkamp et al. 2010). The Balanced Inventory of Desirable

Responding scale measures both components of the social desirability bias, while the remaining scales are highly correlated with one of its two factors.

The Social Desirability Bias in Marketing Research

Research has yielded conflicting evidence concerning the association of the social desirability bias with scales relevant to marketing. For example, Mick (1996) found systematically biased responses to trait materialism measures (i.e., a social desirability bias), while Richins and Dawson (1992) found no such relationship. In a separate study, Richins (1983) found an association between social desirability bias measures and self-reported aggressiveness.

Using the Marlowe-Crowne scale to detect the social desirability bias, Unger and Kernan (1983) examined the possible confounding of the social desirability bias with a six-dimensional scale of leisure. Although no correlation coefficients or significance levels were reported, the authors indicated that there was evidence of situational specific social desirability biases. Unfortunately, given the infrequent use of social desirability bias scales and the overreliance of the unidimensional Marlowe-Crowne scale, evidence of the bias‘ influence on marketing relevant constructs is limited. Yet, research published in psychology has reported relationships between social desirability bias scales and a number of marketing relevant phenomena including attitudes

(Lenski and Leggett 1960), attitude change (Bagozzi and Schnedlitz 1985; Buckhout 1965;

Goldsmith 1987), consumer satisfaction (Sabourin et al. 1989), dyadic interactions (Buckhout

1965), innovativeness (Goldsmith 1987), and risk-taking behavior (Kogan and Wallach 1964).

Consequently, administering measures of the social desirability bias during the construct validation process has become common practice for psychologists (King and Bruner 2000).

Unfortunately, marketing researchers have yet to adopt similar practices. In fact, only thirteen papers, published from 1980 to 1997, across six of marketing‘s more prestigious journals, explicitly measured the social desirability bias (King and Bruner 2000). As an extension, we compiled a database of articles published in the same six marketing journals

(Journal of Marketing Research, Journal of Marketing, Journal of Consumer Research, Journal of the Academy of Marketing Science, Journal of Advertising, and Journal of Advertising

Research) between January 1998 and April 2011 using the keyword social desirability. We evaluated the 88 articles retrieved and found that just 16 articles measured the bias; we present a summary of articles in Table 1. Integrating these results with those of King and Bruner (2000), reveals that only 29 of the hundreds of articles published in these journals over the last 30 years

(from 1980-2011) measured the social desirability bias.

In a similar analysis on the frequency of social desirability bias measures reported in marketing journals, Steenkamp et al. (2010) found that the Marlowe-Crowne social desirability scale appeared just 23 times between 1968 and 2008 in the Journal of Marketing Research,

Journal of Marketing, and Journal of Consumer Research– making it the most frequently implemented social desirability bias scale in these marketing journals. Unfortunately, the

Marlowe-Crowne scale confounds the two dimensions of the social desirability bias. This proposes a possible explanation as to why the literature reports conflicting evidence on the association of the social desirability bias with various marketing measures. Portions of the

Balanced Inventory of Desirable Responding scale appeared seven times; however, just one article reported using the entire scale. In sum, our field does not appreciate the seriousness of the social desirability bias — evidenced by the limited use of scales detecting the social desirability bias. Furthermore, seventy-seven percent of articles employing such a scale use the unidimensional Marlowe-Crowne scale. In the current research, we address this issue by implementing a two-factor model of the social desirability bias.

A Procedure for Detecting The Social Desirability Bias

A number of social desirability bias detection techniques have been proposed in the literature (see Nancarrow, Brace, and Wright 2001). The most common is the simple correlational approach (Paulhus 1991). The correlational approach requires participants to complete a measure of the target construct(s) and a measure of the social desirability bias. The bias is reportedly present when the construct measure correlates with the social desirability bias scale. Conversely, a non-significant correlation coefficient suggests that the construct measure does not elicit socially desirable responses (at least under current study conditions).

Unfortunately, the shortcoming of this, as well as with similar correlational approaches, is well documented (see Paulhus 1991). Of primary concern is the failure of correlational techniques to account for shared variance between the two factors of the social desirability bias (namely, self- deception and impression management) and inability to categorize the bias as deliberate or unconscious distortion. Lastly, correlational techniques fail to distinguish substantive from stylistic responding. In sum, it is difficult to reliably interpret the nature of an observed correlational relationship. For instance, it is unclear whether a significant correlation indicates a true, yet exaggerated response or a complete falsification.

Given these limitations Steenkamp et al. (2010) developed a more conclusive approach to detect and categorize the social desirability bias using experimental methods. This approach involves two steps. Step one investigates the presence of the bias, whereas step two examines if the bias is an unconscious self-deception or a deliberate impression management behavior. The procedure for step one is similar to the correlational approach, however instead of calculating correlation coefficients the researcher must regress the bias onto the construct of interest, Y.

Thus, if we assume the widely accepted two-factor model of the social desirability bias, where the social desirability bias consists of impression management (IM) and self-deception (SDE), step one begins by estimating the following regression model:

Y = b0 + b1IM + b2SDE

According to Steenkamp et al., (2010) a standardized b1 or b2 exceeding |.2| indicates a nonnegligible relationship between Y and one or more of the social desirability bias factors.

Because impression management and self-deception are positively correlated, a standardized coefficient of |.2| corresponds to a zero-order correlation that is halfway between a small and medium effect size. Step two of the substance versus style approach requires that participants be randomly assigned to either a high or low demand condition. When demand for favorable self- presentation is low (vs. high), participant responses exhibiting the social desirability bias are either accurate or distorted unconsciously (vs. deliberately). The following section reviews the literature on indirect questioning, a common technique used to avoid the social desirability bias, and its relationship with construal level theory.

Indirect Questioning & Construal Level Theory

As discussed previously, indirect questioning is a common technique used to attenuate the social desirability bias. According to Fisher (1993), indirect questioning is a structured projective measurement technique that asks respondents to answer questions from the perspective of another person or group (Anderson 1978; Calder and Burnkrant 1977; Robertson and Joselyn 1974). An indirect question may ask the respondent how ―most people,‖ ―the average person,‖ ―the typical person,‖ or ―a friend‖ would respond to a given question or set of items (Alpert 1971; Fisher 1993; Luchs et al. 2010). Using an ambiguous third party allows respondents to project their own beliefs and attitudes when responding (Naylor, Lamberton, and

Norton 2011). In addition, increasing the social distance of the target allows respondents to remain disassociated from the response and its hazardous social implications. In sum, the indirect nature of the question format reduces the social implications of providing an undesirable response.

A distinct line of research on construal level theory proposes that increased social distance has important implications on how we construe events and objects. According to construal level theory, we transcend the here and now by forming abstract mental construals of distal objects (Liberman and Trope 1998; Trope and Liberman 2003; Trope and Liberman 2010).

This mental construal process is essential to recalling the past, empathizing with others, imagining what could have been, and visualizing future events. Each of these mental processes requires a detachment from the self, as it exists here and now (referred to as psychological distance). Thus, psychological distance is egocentric (with the self as the reference point) and the psychological distance of an event or object increases as we move away from this point of origin along any of the following dimensions –time, social space, spatial distance, or probability.

According to construal level theory, we construe psychologically distant objects more abstractly than objects that are psychologically near. In other words, as the psychological distance of an object or event increases, we refocus from detailed, incidental features to central, fundamental characteristics.

Consumers frequently make decisions that involve one or more dimensions of psychological distance (Kim, Zhang, and Li 2008). For example, when contemplating a potential vacation a consumer is likely to consider all of the psychological distance dimensions: temporal distance (e.g., to travel tomorrow vs. next year), spatial distance (e.g., to a destination 10 vs.

2,000 miles away), social distance (e.g., to travel with others vs. alone), and probability (e.g., an infinitesimal vs. certain probability of vacationing). A consumer considering a last minute vacation set to depart tomorrow will focus on concrete (low construal level), task-oriented features centered on how to complete the trip (e.g., traveling to the airport, packing, etc.).

Conversely, if the trip were to depart a year from now attention would shift to abstract, (high construal level) goal-oriented features centered on why to complete the trip (e.g., to have fun, to get away from it all, etc.). Thus, variations in psychological distances alter how a consumer construes the same object or event. In addition, psychological distances influence the utility consumers place on various product attributes (Khan, Zhu, and Kalra 2011).

Researchers have examined the effects of specific psychological distance dimensions on consumer decision making, choice, and inference formation (see Liberman, Trope, and Wakslak

2007). Temporal distance affects the inferences we draw about others and ourselves. For example, Pronin and Ross (2006) showed that compared to the present self, a person is more likely to view their future and past selves in terms of dispositional traits. In a related study,

Wakslak, et al. (2008) asked participants to imagine themselves in various situations in the near

(vs. distant) future and to predict the extent to which they would act in accordance with their own dispositional traits. They found that participants predicted more personality consistent behaviors in the distant (vs. near) future. These findings suggest that people rely on abstract, decontextualized traits when forming temporally distant self-representations. We are unaware of research exploring self-representations along the remaining dimensions of psychological distance, however, the psychological distance dimensions share a positive association, suggesting a common underlying factor (Bar-Anan et al. 2007). These findings suggest that adopting a distal versus a proximal perspective changes the emphasis people place on the incidental aspects of a given situation. Related research on the actor-observer effect (also referred to as correspondence bias), finds that a person‘s attribution of his or her own behavior tends to emphasize contextual factors. On the other hand, observed behavior is often attributed to the dispositional traits of the third party (Jones and Nisbett 1972). Similarly, memories recalled from a third-person perspective, rather than a first person perspective, tend to emphasize dispositional (vs. situational) representations (Nigro and Neisser 1983). Thus, a third-person perspective increases social distance beyond that of a first-person perspective, inducing a higher level of construal (Pronin and Ross 2006). In addition to social distance, the tendency to undervalue situational factors and overvalue trait inferences increases with temporal (Nussbaum,

Trope, and Liberman 2003) and spatial (Fujita et al. 2006) distance (Rim, Uleman, and Trope

2009).

As psychological distance or construal levels increase, values related to the self-concept should take precedence over incidental factors. In accordance with these findings we hypothesize that consumer responses will reflect dispositional traits when the social, spatial, or hypothetical dimensions of psychological distance are distal. Therefore, techniques increasing psychological

12 distance or construal levels should reduce the social desirability bias. To date, indirect questioning has exclusively manipulated social distance to avoid the social desirability bias, however given the association between the psychological dimensions we hypothesize that manipulations of time, space, or probability should also influence this bias. To test our theory we first identify marketing constructs suspected of eliciting the social desirability bias.

Marketing Constructs Eliciting or Suspected to Elicit the Social Desirability Bias

According to King and Bruner (2000), the social desirability bias is of particular concern when self-report measures are used (Zerbe and Paulhus 1987), when measuring sensitive constructs, when anonymity is compromised, or when subjects anticipate responses will result in normatively influenced or evaluated. As a result, we review the relevant literature reporting relationships between self-report measures of marketing constructs and the social desirability bias.

Materialism and Impulsive Buying Tendency. In a large mail survey of 2,500 US households, Rindfleisch, Wong, and Burroughs (2006) found significant correlations between a reduced form of Paulhus‘ (1991) Balanced Inventory of Desirable Responding scale and measures of materialism, personal insecurity, social insecurity, developmental insecurity, existential insecurity, age, gender, and brand loyalty (N = 314). Unfortunately, they did not separate the social desirability bias scale into its two factors (impression management and self- deception); however, the correlation coefficient reported between the social desirability bias scale and materialism (r = -0.37; p < 0.01) was significant. In a separate study, Mick (1996) explored the presence of the social desirability bias in self-report measures of materialism, self-

13 esteem, compulsive buying behaviors, and impulsive buying behaviors. Mick concluded that respondents underreport their levels of materialism value, compulsive buying, impulse buying, and neuroticism and over report their level of self-esteem. Although the authors asserted that the social desirability bias had no significant effect on understanding and explaining the relationships between these variables, they reported significant correlations, suggesting an association between these constructs and the social desirability bias. However, given that the correlational technique was used to detect the social desirability bias we are unable to speculate further on the nature of these associations.

Green Orientation and Corporate Social Responsibility. Corporate social responsibility is a person‘s belief about the ability of companies to produce and deliver quality goods and services while also being socially responsible. Perhaps given its social responsibility component, significant correlations between impression management and various forms of corporate social responsibility (r = -0.28 to .17) have been reported (Hegarty and Sims 1978).

Sustainability has become a critical concern for consumers and marketers. Recent research on the sustainability liability effect demonstrates that consumers are less likely to select sustainable products when efficacy is of primary concern (Luchs et al. 2010). The authors concluded that a negative association between perceived efficacy and sustainability explains this effect. Moreover, they report that social distance moderates this effect, presumably due by attenuating the social desirability bias (Study 2 and Study 5).

Frequency of Healthy Preventative Behaviors, Television Viewing Frequency, and Regulation of

Alcohol Consumption. Researchers have detected the social desirability bias in behavioral

14 as well as dispositional measures. For example, previous research has identified an association between self-reported health conscientiousness and response biases (Baumgartner and

Steenkamp 2001). The social desirability bias is also associated with self-report measures of alcohol consumption (Blair et al. 1977) and television viewing frequency (Phillips and Clancy

1972).

Study 1 Hypotheses

Our investigation begins by identifying marketing constructs eliciting the social desirability bias through Steenkamp et al.‘s (2010) substance versus style approach. Once identified, we continue our investigation by testing if questioning techniques manipulating the various dimensions of psychological distance attenuate the bias. The marketing constructs discussed in the previous section have been found or are suspected to elicit the social desirability bias. Given these previous findings, we predict the following:

H1: Direct questioning should elicit the social desirability bias. More specifically, we

predict that direct (vs. indirect) questioning will result in a non-negligible relationship

with one or both of the social desirability bias factors.

As previously discussed, adopting a third-person perspective increases social distance and construal level. Given that increased psychological distance is associated with trait (vs. situational) inferences we predict the following:

H2: Indirect (vs. direct) questioning increasing social distance should decrease the social

desirability bias.

Given the common factor underlying the various psychological distance dimensions, we propose the same effect for indirect questioning increasing spatial distance (Bar-Anan et al. 2007):

H3: Indirect (vs. direct) questioning increasing spatial distance should decrease the social

desirability bias.

Unlike with social and spatial distance, Trope and Liberman (2010) hypothesized temporal distance would increase positivity (i.e., people will be more positive about the distant future). This hypothesis is congruent with research on the future optimism effect, finding that individuals view their future selves and experiences with confidence and hopefulness (e.g.,

Regan, Snyder, and Kassin 1995; Weinstein 1980). Therefore, we anticipate that:

H4: Indirect (vs. direct) questioning increasing temporal distance should increase the

social desirability bias.

CHAPTER THREE

Study 1

In Study 1, we investigate whether there is an association between the direct measures of seven sensitive marketing constructs and the social desirability bias. In addition, we determine if indirect measurement techniques, manipulating the various dimensions of psychological distance, attenuate the social desirability bias. To detect the presence of the bias, we utilize the substance versus style approach developed by Steenkamp et al. (2010). In accordance with this approach, we manipulate self-presentation concern through perceived anonymity. Although we predict that the mere sensitivity of these self-report measures will be sufficient to elicit the social desirability bias, we manipulate self-presentation concern to further enhance the conscious distortion of responses. In sum, given the sensitive nature of these self-report measures, we predict that we will detect socially desirable responses across both self-presentation concern conditions [H1].

We also manipulate the framing of each question according to the three psychological distance dimensions previously mentioned. These dimensions include spatial, social, and temporal distance. Bar-Anan, Liberman, Trope, and Algom (2007) show that these psychological distance dimensions share a common association, and predict that this association is nonconscious. In a series of studies, they found that the activation of one distance dimension automatically activated the perceived distances along the remaining dimensions. Similarly, a growing body of work, manipulating the various dimensions of psychological distance, demonstrates similar effects across dimensions (e.g. Amit, Algom, and Trope 2009; Stephan,

Liberman, and Trope 2010).

However, in this study, the psychological distance manipulations used possess an egocentric reference point (i.e., the participant). Participants respond from a distant future self or location perspective, or from the perspective of another person. Our main prediction is that by increasing psychological distance, we deemphasize contextual demand biases to engage in the social desirability bias. With an egocentric reference point, participants project their own attitudes and behaviors on to each response, thus providing accurate responses that are unaffected by demand biases. Thus, we predict that indirect measures manipulating social [H2] and spatial distance [H3] will attenuate the social desirability bias. However, as previously discussed, asking individuals to respond from a future self perspective will increase the likelihood that they will provide distorted responses given the future optimism effect (e.g., Regan et al. 1995; Weinstein 1980). Therefore, we predict that an indirect measure manipulating temporal distance will increase the social desirability bias [H4].

Method

A 3 (Self-presentation concern: High or Low or Control) x 4 (Questioning method:

Indirect Using Temporal Distance or Indirect Using Spatial Distance or Indirect Using Social

Distance or Direct) between subjects design was used. Questioning method was manipulated to test our hypotheses, while self-presentation concern was manipulated to elicit and categorize the observed social desirability bias as either deliberate impression management or unconscious self- deception. The dependent variables of interest were measures of dispositional (i.e., Materialism,

Impulsive Buying Tendency, CSR, and Green Orientation) and behavioral (i.e., Frequency of

Healthy Preventative Behaviors, Television Viewing Frequency, and Regulation of Alcohol

Consumption) constructs suspected to elicit the social desirability bias. Manipulations and measures for Study 1 can be found in Appendix A.

Materialism. Materialism was measured using the nine-item, Material Values Scale used by

Rindfleisch et al. (2006) and Richins (2004). This scale assesses the degree to which an individual views the acquisition of material objects as an important life value. The Material

Values Scale has been employed in dozens of prior applications. After participants completed the scale, a materialism composite measure was created by averaging the scores (α = .81).

Impulsive Buying Tendency. Impulsive buying tendency was measured using the eight-item scale developed by Martin, Weun, and Beatty (1994) and employed by Mick (1996). After participants completed the impulsive buying tendency scale, a composite measure was created by averaging the scores (α = .71).

Corporate Social Reasonability. Corporate social reasonability was measured using Sen and

Bhattacharya‘s (2001) nine-item scale. This scale measures a person‘s beliefs about the ability of companies to produce and deliver quality goods and services while also being socially responsible. After participants completed the corporate social reasonability scale, a composite measure was created by averaging the scores (α = .86).

Green Orientation. We used Dunlap and Van Liere‘s (1978) twelve-item scale (also referred to as the New Environmental Paradigm or Environmentalism scale) to assess Green Orientation.

This scale measures a respondent‘s attitude towards and perspective on humanities‘ relationship

19 with the environment. After participants completed the green orientation scale, a composite measure was created by averaging the scores (α = .77).

Frequency of Healthy Preventative Behaviors. A scale measuring the extent to which respondents engage in five activities related to maintaining good health was measured (Jayanti and Burns 1998; Moorman and Matulich 1993). The activities included eating a well-balanced diet and reducing the amount of salt content in their diet. After participants completed the scale, a composite measure was created by averaging the scores (α = .8).

Television Viewing Frequency and Regulation of Alcohol Consumption. The regularity of television viewing and regulation of alcohol consumption were both measured with single items.

Each item asked participants to indicate the frequency with which they engage in the respective activity on a scale ranging from 1 (―never‖) to 7 (―always‖).

The Social Desirability Bias. The social desirability bias was measured using the 20-item version of the Balanced Inventory of Desirable Responding (BIRD) scale developed by Paulhus

(1991) and refined by Steenkamp et al. (2010). The first 10 items of the Balanced Inventory of

Desirable Responding scale measure self-deception, which is a conscious tendency to view oneself favorably, while the remaining items measure impression management, which is the conscious distortion of responses to impart favorable impressions. The complete scale can be found in the Appendix D. After participants completed the scale, self-deception (α = .77) and impression management (α = .71) composite measures were created by averaging the within factor scores.

Perceived Anonymity. Perceptions of anonymity were measured using a two-item scale.

Participants indicated how strongly they agree with the following statements ―The responses I provide on this survey are anonymous‖ and ―The responses I have provided will remain confidential‖ on a seven-point scale ranging from 1 (―strongly disagree‖) to 7 (―strongly agree‖).

The complete scale can be found in the Appendix D. After participants completed the scale, a composite measure was created by averaging the scores (α = .62).

Four hundred and fifty-three subjects were sampled from the online subject pool Amazon

Mechanical Turk and paid for their participation. Participants were screened based on prior approval rating (the percentage of prior tasks they had completed that had been approved by the task issuing party was 95% or greater) and age (minimum age was 18 years); see Table 2 for a sample description. Participants were given a link to the experiment, which was administered using the online survey tool Qualtrics. The participants were randomly assigned to experimental conditions. To manipulate self-presentation concern, participants were told that the responses they provide in the forthcoming experiment would either be (i) kept fully confidential

(anonymous condition), (ii) posted online along with identifying information (public condition), or (iii) given no information (control condition). After being presented with the self-presentation concern condition, participants were asked if they would like to proceed. Participants responding

―yes‖ proceeded to the following section, whereas participants responding ―no‖ voluntarily withdrew from the experiment. In all, twenty-five participants voluntarily withdrew from the experiment (80% from the public condition).

Participants were then randomly assigned to one of four question format conditions.

Participants assigned to the direct question format condition were presented with the original version of each marketing scale (i.e., with no psychological distance manipulations). Participants assigned to one of the remaining question format conditions were asked to complete the same seven scales with embedded temporal, spatial, or social distance manipulations. Temporal distance was manipulated by asking participants to predict how they would respond to the scale a year into the future. Whereas, spatial distance was manipulated by asking participants to imagine they were in another area, far from where they are now and with this in mind how they respond to the given questions. Lastly, social distance was manipulated by asking subjects to predict the likely responses of ―most people.‖ Prior research has established the effectiveness of similar manipulations in increasing psychological distance (Chandran and Menon 2004; Liberman and

Trope 1998). The scales were randomly presented across all conditions. This was followed by measures of the social desirability bias, perceived anonymity, and relevant demographic variables. Before completing the survey, participants were debriefed as to the true purpose of the study and reassured that all responses would remain fully confidential.

Results

In accordance with the substance versus style approach, we investigated the potential of a social desirability bias by performing a series of regression analyses. To investigate differences in the occurrence of the social desirability bias, we regressed self-deception and impression management onto each marketing scale, and within each experimental condition. Figure 1 plots the standardized regression coefficients across conditions and scales.

The plots reveal conditions under which the seven marketing scales exceed the acceptable standardized |.2| cutoff for self-deception and impression management set by Steenkamp et al.

(2010). To test our hypotheses we initially compared the standardized coefficients within the self-presentation control condition only, as outlined by Steenkamp et al. (2010). Overall, the results support our hypotheses. When using direct questioning, five of the seven constructs exhibited a non-negligible relationships with one of the social desirability bias‘ two factors (i.e.,

Materialism; Green Orientation; Frequency of Healthy Preventative Behaviors; Television

Viewing Frequency; and Regulation of Alcohol Consumption). Only impulsive buying tendency and corporate social responsibility produced standardized beta coefficients within the acceptable range, thus supporting H1. We also hypothesized that indirect questioning manipulating social or spatial distance would attenuate the social desirability bias, H2 and H3 respectively. Results provide support to H2. Only two of the five factors that demonstrated a relationship with the social desirability bias using direct questioning displayed a non-negligible relationship using indirect questioning increasing social distance (i.e., Frequency of Healthy Preventative

Behaviors; and Television Viewing Frequency).

In support of H3 only one of the five constructs exhibited the social desirability bias under conditions of indirect questioning increasing spatial distance (i.e., Frequency of Healthy

Preventative Behaviors). Thus, we conclude that spatial or social distance reduces the social desirability bias. The results also lend support to H4, given that all five constructs exhibited the social desirability bias when the question format manipulated temporal distance. Our results also lend support to the reliability of the materialism scale, as the standardized coefficient for impression management remained within the acceptable range of |.2| across all conditions.

Discussion

In Study 1, we find evidence, across seven marketing constructs, that indirect questioning increasing social or spatial distance reduces the social desirability bias as compared with direct questioning. We also find evidence that indirect questioning manipulating temporal (i.e., future) distance increases the social desirability bias. We expected this result given research on the future optimism effect. These results lend support to construal level theory as a theoretical explanation for indirect measurement techniques. Question formats that increase social and spatial distance increase psychological distance and construal levels, thus deemphasizing contextualized cognitions (i.e., demand biases), while emphasizing dispositional cognitions (i.e., traits). In Study 2, we extend these findings to a consumer choice context. In doing so, we test if the effects observed in Study 1 influence consumer choice behavior. We also discuss the association between construal level and psychological distance and test if construal levels directly influence the social desirability bias.

Do the observed effects represent unconscious self-deception or deliberate impression management? In accordance with step two of the substance versus style approach, we randomly assigned participants to a high or low self-presentation concern condition. Posttest manipulation checks verified that participants in the anonymous condition perceived that their responses would remain anonymous (M = 5.19, SD = 1.37) as compared with participants in the public condition

(M = 4.66, SD = 1.41; t(1, 259) = 3.13, p < .01). We interpret this as evidence of successful manipulation of perceived anonymity.

According to Steenkamp et al. (2010) when the demand for favorable self-presentation is low (vs. high), the self-reports provided by participants exhibiting the social desirability bias are either accurate or distorted unconsciously (vs. deliberately). To categorize the bias the

24 distribution of scores for the sensitive marketing constructs for individuals high in the social desirability bias tendency should be compared across demand conditions. If the scores are higher

(vs. lower) for socially desirable (vs. undesirable) characteristics in the high (vs. low) demand condition than the social desirability bias is likely deliberate.

To identify the observed relationship with the social desirability bias as a deliberate distortion, we estimated the effect of anonymity (0 = anonymous; 1 = public) on each of the seven marketing scales, while setting each dimension of the social desirability bias to 1 standard deviation above the mean. In all, we estimated 14 regression models, one for each marketing scale and for each dimension of the social desirability bias (i.e., self-deception and impression management). The regression analyses were restricted to participants in the direct questioning condition only (n = 74). In sum, there were no significant differences in the estimated mean values across conditions or measures (ps >.1). Although this analysis does not preclude the possibility of unconscious self-deception or stylistic responding, conscious distortion is unlikely.

We return to this issue in Study 3, by investigating the underlying mechanism. We find similar results in Study 3. We observe a significant in-direct effect between self-presentation-demand, psychological distance, and their interaction on consumer responses, but do not observe a direct effect.

To support our hypothesized theoretical model, it is important to establish a direct relationship between construal level theory and the social desirability bias before testing a full mediation model. Therefore, in Study 2, we utilize a construal level mind-set manipulation to determine if construal levels (vs. psychological distances) influence the social desirability bias.

By manipulating construal levels in an unrelated task, we test our hypothesis that construal level decreases the social desirability bias.

CHAPTER FOUR

Study 2

In Study 1, we demonstrate that indirect questioning manipulating the framing of a question in terms of time, space, and social distance influenced the association of social desirability bias scales with seven marketing constructs. However, as suggested by Mick (1996),

―even when contamination is suggested through simple correlations, its influence on the assessment of relationships in a nomological net may be small or nonexistent. Only the systematic analyses of spuriousness, suppression, and moderation effects can address these issues.‖ Thus, in Study 2, we extend our findings by exploring how our construal level theory accounts for the social desirability bias within a consumer choice context. More specifically, we examine how construal level influences the share of choice between sustainable and non- sustainable product options.

According to Luchs et al. (2010), although consumers intend to purchase and use eco- friendly products, they fail to do so because of negative performance inferences (i.e., the sustainability liability effect). Sustainable product attributes are associated with ‗gentleness‘ and

‗tenderness,‘ which negatively affects quality and performance inferences. Thus, explicit quality cues attenuate the negative effect of sustainability on perceived quality by impeding the formation of quality inferences.

Consumers construe sustainable actions, products, and attitudes as pro-social in nature

(i.e., beneficial to society), as a result, there exists a social pressure to display attitudes and behaviors endorsing sustainability, especially when in the presence of others. To demonstrate this demand bias Luchs et al. (2010) asked participants to evaluate sustainable and non- sustainable consumer products while manipulating the ‗point of view‘ of each question. When

26 the question referred to the participant directly, respondents displayed attitudes supporting the socially preferred sustainable product, however when the question referred to a third party (i.e., the average American consumer) the sustainability liability effect emerged, where participants preferred the less sustainable, but better performing, product option. When social distance was low, participants were directly linked to each response, thus increasing the likelihood that they would provide systematically biased responses favoring the socially acceptable standard (i.e., the sustainable product option). However, when social distance was high, responses were no longer directly linked to each participant, thus allowing respondents to favor the product that would maximize functional utility (i.e., the less sustainable product). Luchs et al. (2010) observed the sustainability liability effect for attitudes and behaviors across numerous product categories and settings. The following section discusses the relationship between psychological distance and construal level before introducing Study 2.

Previous research shows that psychological distance has a significant influence on how we construe objects and events (see Trope and Liberman 2010). As psychological distance increases (according to time, space, social distance, and probability) consumers are more likely to represent psychologically distant objects and events in high-level, abstract, and global terms that emphasize the fundamental properties of an object or event (e.g., superordinate categories, key attributes, primary reasons for performing an activity). Conversely, as distance on any of these psychological distance dimensions decreases, consumers are more likely to represent objects and events in low-level, concrete, and local terms that emphasize the peripheral issues of an object or event (e.g., subordinate categories, peripheral attributes, secondary reasons for performing an activity). Furthermore, the relationship between the psychological distance dimensions and construal level is bidirectional, such that as construal levels increase, inferred

27 psychological distances increase, and vice versa (Trope and Liberman 2010). Thus, one would expect the influence of demand biases, such as the desire to engage in impression management behavior to display attitudes and behaviors consistent with social norms, to decrease as construal levels increase. Therefore, we hypothesize the following regarding construal level and consumer share of choice between a sustainable and non-sustainable product:

H5: When construal level is high choice probabilities for non-sustainable (vs.

sustainable) products will increase relative to when construal level is low.

Method

Undergraduates from a large Midwestern university (N=98) participated in exchange for course credit. Participants were randomly assigned to either a high construal level mind-set condition (n=49) or a low construal level mind-set condition (n=49). The mind-set tasks were identical to those used by Freitas et al. (2004), except they were administered online. In both conditions, participants were asked to consider a prompt involving improving and maintaining one‘s physical health. In the high-level mind-set condition, participants responded to a series of questions directing them to consider why it might be important to improve and maintain one‘s physical health, while in the low-level mind-set condition, participants were directed to think about how one might improve and maintain one‘s health. Prior research demonstrates the effectiveness of this manipulation in directly influencing construal levels (Agrawal and Wan

2009; Wakslak and Trope 2009; Wright et al. 2011; Yan and Sengupta 2011).

After the mind-set manipulation task, participants chose between two laundry detergents that were on the market and available at retail locations accessible to the sample population during the data collection period. The first laundry detergent, manufactured by Seventh

Generation, was positioned as the sustainable option with the slogan ―Tough on dirt but gentle on the environment‖ and given environmentally conscientious product features such as ―the

Laundry Liquid bottle is made from 80% post-consumer recycled plastic‖ and ―biodegradable formula.‖ The second laundry detergent, manufactured by Tide, was positioned as the non- sustainable option with the slogan ―for a brilliant clean every time‖ and given performance centric product features such as ―includes a clinically proven, anti-bacterial formula‖ and

―designed for stain removal.‖ Participants chose between these two randomly presented choices prior to completing a series of scales asking participants to indicate which detergent is better for the environment, better at cleaning clothes, and more gentle, 1 (definitely Tide) to 11 (definitely

Seventh Generation). Manipulations and measures for Study 2 can be found in the Appendix.

Results

We asked participants which detergent is better for the environment, better at cleaning clothes, and more gentle, 1 (definitely Tide) to 11 (definitely Seventh Generation). Participants correctly identified Seventh Generation as being better for the environment (M = 9.59, SD =

2.16, the mean value significantly differs from the scale midpoint of six; t(97) = 16.44, p < .001).

While both detergents were considered to be equally gentle (M= 6.34, SD = 2.87, the mean value did not significantly differ from the scale midpoint of six, p > .1), Tide was rated as being better at cleaning clothing (M = 3.37, SD = 1.96, the mean value significantly differs from the scale midpoint of six; t(97) =-13.29, p < .001).

A logistic regression showed that construal level significantly influenced choice (χ2(1,97)

= 4.48, p < .05). As expected, when participants were in a low-level (vs. high-level) construal level mind-set they were more likely to choose the sustainable (vs. non-sustainable) laundry

29 detergent (27% versus 10%). Thus, when construal level was high the demand to provide a socially acceptable response (i.e., to choose Seventh Generation) was attenuated, thus allowing respondents to select the better performing product (i.e., Tide). Luchs et al. (2010) demonstrated the same effect using traditional indirect questioning, whereby they increased the social distance of the question target. Study 2 extends the work of Luchs et al., along with the results in Study 1 by demonstrating that construal levels rather than psychological distances reveal the sustainability liability effect. We find that increasing construal levels reveals actual preferences versus those imposed by contextual demands. Using construal level mind-set manipulations (vs. psychological distance dimensions) provides further evidence of the relationship between construal level theory and the social desirability bias. Study 3 further explores the mechanism behind this relationship while extending the effect to a novel product category.

Discussion

Study 2 results demonstrate that compared to consumers in a low-level mind-set, consumers in a high-level mind-set were more likely to select the non-sustainable product option.

These results are consistent with our hypothesized relationship between construal levels and social desirability responding. As demonstrated by Liberman and Trope (1998), increasing construal levels deemphasizes contextual details. We posit that as construal levels increase, the prominence of contextually dependent demand biases, such as the need to engage in impression management behavior, become deemphasized, thus resulting in consumer responds unaffected by contextual demands.

However, past research suggests an alternative explanation that may account for our observed effects. In a study focusing on the relationship between price and quality, Yan and

Sengupta (2011) concluded that compared to specific, concrete product attributes, an attribute like price can be thought of as a more abstract, general cue, especially for quality inferences.

They further explained that one possible explanation for the relatively more abstract nature of price, compared with other attributes, is that price is universal to practically all products and that consumers use price as a general quality cue. In a series of experiments, manipulating psychological distance dimensions and construal levels, the authors concluded that given high- level construals, a reliance on abstract cues, such as price, increases. Whereas, given low-level construals, a reliance on product-specific attributes increases.

Similar to price, consumers may construe a product‘s sustainability as being an abstract attribute. As reflected in studies on willingness to pay, consumers view the purchasing of sustainable products as beneficial to society (Trudel and Cotte 2008). Therefore, purchasing a non-sustainable product may negatively influence a broad spectrum of humanity, over a long period. Consequently, consumers are willing to pay more for products that are sustainable, and as observed with price, associate sustainability with higher quality. If consumers perceive the sustainability of a product as an abstract attribute, then its influence on choice should increase as construal levels increase. In Study 3, we address this alternative account by testing the meditational process explaining the association between construal level theory and the social desirability bias. In addition, we return to psychological distance as our manipulation of construal levels and test our mediation model across multiple consumer choices.

In sum, Study 3 has three primary objectives. First, we directly measure the social desirability bias to investigate the underlying process explaining the effect of construal level on consumer choice behavior. Second, to test the generalizability of our results we examine our hypothesized model across multiple product categories. Lastly, in accordance with prior research

31 on construal level theory demonstrating the bidirectionality of construal levels and psychological distances (Trope and Liberman 2010), we manipulate construal levels through the psychological distance dimensions.

CHAPTER FIVE

Study 3

In Study 2, we provide initial evidence that increasing construal levels alleviates the demand to yield to socially preferable choice behavior, thus revealing actual consumer preferences. In Study 3, we extend these findings in the following three ways. First, we return to psychological distance as our projective technique. Doing so allows us to further test construal level theory as a theoretical account explaining and extending projective measurement techniques. Second, we measure the social desirability bias to test and categorize the meditational process as a deliberate distortion. Third, we test the generalizability of our results by incorporating multiple product categories.

To elicit impression management behavior we manipulate the presence of a self- presentation concern. In addition, we manipulate the psychological distance of the choice task itself to reduce conscious distortion. Prior research shows that social desirability bias scales are sensitive to demand manipulations, suggesting they capture deliberate distortion (McFarland and

Ryan 2006; Pauls and Crost 2004; Roth and Herzberg 2007). Hence, asking respondents to provide favorable responses (i.e., to ―fake good‖) should encourage deliberate misrepresentation.

This high demand condition provides a rigorous test of our theory that increasing psychological distance should decrease deliberate impression management behavior. Thus, we anticipate the following:

H6: When self-presentational concerns are high, impression management will be higher

when psychological distance is low, compared to when psychological distance is high.

In addition to measuring the two dimensions of the social desirability bias (i.e., impression management and self-deception), we will observe consumer choices between products that differ solely according to their sustainability. Our main prediction is that the interactive effect of the demand for self-presentation and the psychological distance of the choice task on choice will be mediated by the conscious dimension of social desirability responding

(i.e., impression management). Thus, we anticipate the following mediated moderation effect:

H7: Impression management will mediate the effect of self-presentation concern and the

psychological distance of the choice task on choice. (interaction  impression

management  choice).

Method

A 2 (Self-presentation concern: high vs. low) x 2 (psychological distance of choice: proximal vs. distal) between-subjects design was used. Similar to Study 2, participants were presented with choices between sustainable and non-sustainable products. The primary variables of interest were the two dimensions of social desirability responding and selection of either the sustainable or non-sustainable product. The psychological distance of the choice task was either psychologically proximal or distal. Manipulations and measures of Study 3 can be found in

Appendix C.

Self Presentation Concern. Self presentation concern was manipulated by instructing participants to present themselves as ―favorably as possible,‖ while completing the study (see Paulhus 2002), versus a control condition. Several studies have demonstrated the effectiveness of this

34 manipulation in increasing impression management behavior (Blake et al. 2006; Paulhus and

Reynolds 1995).

Psychological Distance of Choice. The psychological distance of the choices were manipulated through the four dimensions of psychological distance (i.e., time, space, social distance, and probability). Thus, participants in the psychologically proximal condition were presented with the following choice task, ―Assuming that you are likely to make a purchase, which product would you choose for yourself to use here and now?‖ While participants in the psychologically distal condition received the following choice task, ―Assuming that you are unlikely to make a purchase, which product would you choose for an acquaintance to use somewhere else at another time?‖ Therefore, we manipulated the temporal, spatial, social, and probable dimensions of the choice task simultaneously.

The Social Desirability Bias. The social desirability bias was measured using the same 20-item version of the Balanced Inventory of Desirable Responding (BIRD) scale used in Study 1. The first 10 items of the Balanced Inventory of Desirable Responding scale measure self-deception, which is a conscious tendency to view oneself favorably, while the remaining items measure impression management, which reflects a conscious distortion of responses to impart favorable impressions. The complete scale can be found in the Appendix D. After participants completed the scale, the self-deception (α = .62) and impression management (α = .78) indices were created by averaging the within factor scores.

Choice. Participants completed two randomly presented choice tasks, each between two products within the same product category. For each choice, one product was superior to the other in terms of sustainability, but inferior in terms of performance. The product categories included laundry detergents and hand sanitizers. We used the product descriptions described in Study 2 for the laundry detergent category, whereas for hand sanitizers, we manipulated the objective sustainability and performance of the product options. We adapted the manipulation used by

Luchs et al. (2010; Study 2), whereby the sustainability of each product was manipulated using a statement from a fictitious independent agency, the ―Ethical Product Council (EPC).‖ The EPC

―rates similar products based along a variety of proenvironmental and prosocial factors such as sensitivity about pollution and resource usage as well as fair treatment of staff, suppliers, and communities.‖ As Appendix C shows, although the description of the size, uses, and availability of the hand sanitizers remained constant, the Ethical Product Council and performance ratings differed (10 = ―superb‖ versus 2 = ―poor‖). We intentionally positioned one hand sanitizer as being superior in terms of sustainability, but inferior in terms of performance (i.e., superb EPC rating and poor performance rating), while the opposite was true for the other hand sanitizer (i.e., poor EPC rating and superb performance rating).

Manipulating (vs. measuring) a contrast between sustainability and performance provides a rigorous test of the sustainability liability effect and of our hypotheses. We expect participants to select the sustainably inferior, yet performance superior, option when the choice is psychologically distal or when self-presentation concern is low. However, when the choice is psychologically proximal (thus closely associated with the chooser) and when self-presentation concern is high, the pressure to maintain a positive impression will increase the likelihood that consumers will choose the sustainable product. To test for the mediating effect of impression

36 management (i.e., a conscious distortion of responses), participants completed measures of the social desirability bias before completing demographic measures and after the choice tasks.

Four hundred and twenty-seven US adults were sampled from the online subject pool

Amazon Mechanical Turk and paid for their participation. Participants were screened based on prior approval ratings (the percentage of prior tasks they had completed that had been approved by the task issuing party was 95% or greater) and age (minimum age was 18 years). Participants were given a link to this experiment, which was administered using the online survey tool

Qualtrics. The participants were randomly assigned to experimental conditions. To manipulate self-presentation concern, we instructed participants to present themselves as favorably as possible while completing the survey or gave no information (low condition). Participants were then asked to make two product choices. The product options and categories were randomly presented. As previously described, the psychological distance of the choices were either psychologically proximal or distal.

Results

The Meditational Role of Impression Management on Choice Behavior. We propose that the interaction between self-presentation concern and the psychological distance of the choice between two products that differ according to sustainability is mediated by impression management. We estimated a multiple mediator model for each choice with the interaction between self-presentation concern and the psychological proximity of the choice as the independent variable (+1 = high self-presentation concern; and low psychological distance; -1 =

37 all other conditions), and, the mean-centered values for impression management and self- deception as our proposed mediators, and the main effects of self-presentation concern and the psychological proximity of the choice as covariates. Both impression management and self- deception were simultaneously included in the multiple mediator models, thus reducing the likelihood of parameter bias due to omitted variables (Preacher and Hayes 2008) and to determine if the observed effect is explained by a conscious or unconscious process. Significance levels were based on bias-corrected (BC) bootstrap confidence intervals (CIs).

First, we discuss the relationship between our independent variable and our mediating variables, our analysis revealed a significant effect between impression management and the interaction term (β = .27, p < .006). As expected, impression management increased when the choice was psychologically proximal and when self-presentation concern was present. Second we discuss the relationship between our mediating variables and our dependent variable choice.

Our analysis revealed a significant relationship between impression management and the choice between laundry detergents (β = .28, Wald χ2 = 6.19, p < .02) and hand sanitizers (β = .28, Wald

χ2 = 4.50, p < .04). As expected, impression management increased the share of choice for the more (vs. less) sustainable product option. From our meditational analysis, we conclude that the social desirability bias is a conscious distortion (i.e., impression management) and not self- deception (ps > .1).

Table 3 presents the standardized total, direct, aggregate indirect and specific indirect effects of impression management demand, psychological proximity of the choice, and their interaction term on choice via impression management and self-deception. To test the significance of the specific indirect effects we used Preacher and Hayes‘ (2008) multiple mediator technique. This analysis technique estimates bias-corrected (BC) and bootstrap

38 confidence intervals for each indirect effect and allows for the inclusion of both continuous and dichotomous inputs. The results for the overall sample indicate that the effect of our interaction term on the choice between laundry detergents (β = .08, p < .01) and hand sanitizers (β = .08, p <

.05) is fully mediated by impression management. No additional direct or indirect effects were significant (ps > .1). These results support our predicted mediated moderation effect (Baron and

Kenny 1986).

Thus, the results of our analysis support our hypothesized mediated moderation model with impression management as the mediating variable. To further classify our two models, we omitted self-deception, leaving impression management as the sole mediator, and re-ran our hypothesized model. As depicted in Figures 3 and 4, we find support for our mediated moderation model. Given the significant indirect effect of our interaction term on choice through impression management and no significant direct effect we classify our mediated moderation models as indirect-only mediation models (Zhao, Lynch, and Chen 2010).

Discussion

Study 3 demonstrates that when choosing a sustainable product impression management mediates the effect between choice and the interactive effect of self-presentation concern and psychological distance. More specifically, we find that when the choice is psychologically distant impression management decreases, thus revealing actual consumer preference. However, when self-presentation concern is high, and the choice is psychologically proximal, consumers are more likely to make choices that align with social norms. Therefore, within the context of choosing between a sustainable and non-sustainable product, increasing the psychological distance of the choice is an effective means of attenuating social desirability responding and

39 revealing true preferences. This finding provides additional clarity to the underlying mechanism between psychological distance, impression management, and choice.

CHAPTER SIX

General Discussion

Self-report measures assessing socially sensitive constructs are prone to contain numerous biases representing social desirability responding. The social desirability bias systematically biases data, distorting measures of consumer preferences, attitudes, and behaviors.

Despite these clear consequences, marketing researchers continue to pay insufficient attention to the threat posed by this bias (see Fisher 1993; and Steenkamp et al. 2010 for exceptions). Given the time and effort designing a research study and in subsequent data analysis and reporting, the investment in safeguarding against the distortion introduced by the social desirability bias seems quite valuable. Moreover, this bias has important managerial implications. Unmistakably, the social desirability bias weakens the validity of research conclusions. By ignoring this bias, managers relying on the conclusions drawn from biased data risk executing sub-optimal or erroneous business decisions.

Across three studies, we demonstrate the distortion introduced by the social desirability bias. In addition, we develop techniques that attenuate this bias. We propose construal level theory as a conceptual model explaining our techniques. Before developing and testing their effectiveness, we conceptually integrate projective measurement theory with construal level theory. We note the similarity of a classic projective measurement technique (i.e., indirect questioning) with the social distance manipulations common to the construal level theory literature (see Trope and Liberman 2010). To our knowledge we are the first to indentify these similarities and integrate these lines of literature. Next, we review research demonstrating the interchangeable nature of the psychological distance dimensions of space, time, probability, and social distance. Consistent with this research, we propose a series of hypotheses positing the

41 effect that each dimension may have on the social desirability bias. In three experiments, we test our proposed hypotheses and demonstrate how psychological distance and construal levels influence the social desirability bias.

Study 1 provides initial support for our hypotheses. Using the proposed experimental methodology to detect the social desirability bias (see Steenkamp et al. 2010), we tested how questioning techniques, manipulating psychological distances, affected the occurrence of the bias within a survey. Scales measuring socially sensitive constructs directly, exhibited a non- negligible relationship with measures of the social desirability bias; however, this relationship was attenuated when we increased the social or geographic distance of the question. These results provide preliminary evidence supporting our hypothesis that question formats manipulating the psychological distance dimensions proposed by construal level theory attenuate the social desirability bias. However, the particular dimension of psychological distance is an important factor. While social and spatial distance appear to attenuate the association of social desirability bias measures with measures of socially sensitive marketing constructs, we find that asking respondents to refer to a future self increases this association. These results support Trope and Liberman‘s (2010) proposition that taking a future perspective may increase positivity, which consequently explains the observed relationship with the social desirability bias. This conclusion is also consistent with research on the future optimism effect (e.g., Regan et al. 1995;

Weinstein 1980). Future research is needed to explain this effect and its underlying processs(es).

Although these results appear to support our construal level theory framework, they do not identify construal levels as the underlying mechanism explaining the attenuation. We hypothesized that psychological distance would increase construal levels, thus deemphasizing contextual demand biases and allowing respondents to project accurate responses. Although prior

42 research demonstrates the bidirectional relationship between psychological distance and construal levels, we did not test for this relationship. Moreover, our investigation was limited to relationships between marketing and social desirability bias scales. A behavioral outcome variable, shown to elicit the social desirability bias, would provide a more rigorous test of our theory. We addressed these limitations in Study 2.

In Study 2, we provide additional evidence supporting our theoretical model. We reviewed research demonstrating the bidirectional nature of psychological distance and construal level and hypothesized that increasing construal levels directly would decrease the social desirability bias. We tested our hypothesis by inducing participants into either a high or low- level construal mind-set before asking them to choose between a sustainable or non-sustainable product. Prior research on the sustainability liability effect, demonstrates the effectiveness of such choices in eliciting the social desirability bias (Luchs et al. 2010). As previously discussed, consumers are likely to choose a sustainable (vs. non-sustainable) product when the demand to engage in impression management behavior is high (i.e., when observed by a third party). As expected, compared to those in a high-level construal mind-set, respondents in a low-level construal mind-set were nearly three times more likely to select the sustainable production option, thus exhibiting the sustainability liability effect. These results provide additional evidence of the association between projective measurement theory and construal level theory.

Increasing construal levels directly or through psychological distances de-emphasize contextual demand biases, thus, allowing respondents to report accurate responses.

Although these results provide additional support to our construal level theory framework, they do not rule out one possible rival explanation. As a product attribute,

―sustainability,‖ could be construed as an abstract feature. If this account is accurate, according

43 to construal level theory those induced with a high-level construal mind-set should value abstract product features over concrete features and would subsequently choose the sustainable product

(Yan and Sengupta 2011). Thus, the choice behavior demonstrated in Study 2 could be explained by a match between the construal mind-set and the abstractedness of sustainability. We addressed this issue in Study 3, by explicitly testing if the social desirability bias mediates the relationship between psychological distance and the observed choice behavior. We incorporated both dimensions of the social desirability bias to further categorize the bias as a conscious distortion or as self-deception. We also tested the generalizability of our findings by incorporating multiple product categories.

In Study 3, we provide additional clarity into the underlying mechanism by manipulating self-presentation concern and the psychological distance of the choice task while incorporating a measure of the social desirability bias. The results provide evidence of an indirect-only moderated mediation model. When the demand to engage in impression management behavior was high and the psychological distance of the question near, the deliberate distortion (i.e., impression management) component of the social desirability bias explained the observed choice behavior. This effect became non-significant when self-presentation concern was absent or the question format was psychologically distant. We interpret these results in support of our conceptual model. When the demand to engage in impression management behavior is high, increasing the psychological distance of the choice task successfully deters respondents from deliberately providing socially desirable responses. We also report a significant relationship between the impression management dimension of the social desirability bias and choice, whereby the impression management tendency increased the likelihood of selecting the sustainable (vs. non-sustainable) product option. A significant indirect effect between our

44 interaction term and choice through impression management provided evidence of the mediational process and identified the distortion as deliberate. These effects generalized across two choice tasks, product categories, and experimental stimuli.

These results provide additional evidence in support of our construal level theory framework. Establishing impression management as the mediating variable rules out the alternative, feature abstractedness, account for our Study 2 results. Additional research by White et al. (2011), examining the interactive effect of message framing and construal level on recycling behavior, further weakens the likelihood that the feature abstractedness account explains our results. Specifically, these authors demonstrate that the framing and construal level of a message has an interactive effect on recycling intentions and behaviors. More importantly, across three studies they report non-significant main effects of construal levels and psychological distances on recycling intentions and behaviors. Assuming sustainability and recycling are similar in nature, the feature abstractedness account would have predicted main effects in both cases, whereby abstract construal levels or psychologically distant messages would increase recycling intentions and behaviors. In sum, we find converging evidence supporting our proposed model.

It is interesting to note that we did not observe a direct effect between our interaction term and choice. It is well understood that a direct effect between an independent and dependent variable is not necessary for mediation to exist (Zhao et al. 2010). Moreover, some suggest that finding a significant indirect effect can be more powerful than a direct effect because it is a more precise explanation of how the independent variable affects the dependent variable (MacKinnon,

Coxe, and Baraldi 2012). Although one should exercise caution when interpreting non- significant findings, the non-significant direct effect proposes some interesting areas for future

45 research. One possible explanation includes the possibility that an omitted moderating variable, exhibiting competing effects, may have obscured the direct effect. Future research should explore this issue in further detail.

We have shown that increasing construal levels attenuates the social desirability bias.

When demand biases are present, increasing construal levels directly or through psychological distances effectively deters respondents from providing socially desirable responses. We have demonstrated the effectiveness of both techniques within a survey and consumer choice context.

In our choice studies, we presented consumers with a choice between sustainable and non- sustainable product options. Future research should test if increasing construal levels attenuates the social desirability bias in other contexts. For example, Kurt, Inman, and Argo (2011) demonstrate that agency-oriented consumers spend significantly more when shopping with a friend (vs. alone), whereas this effect is attenuated for communion oriented shoppers. While potentially beneficial to the retailer, the social influence unwittingly exerted by the presence of a friend is financially deleterious to the consumer. The authors identify impression management behavior as the underlying mechanism explaining the observed effect. Given our findings, we would expect construal levels to attenuate this effect by diverting the consumers‘ focus away from the contextual demand to manage their friend‘s impression. Such findings would equip consumers with an effective means of averting this social influence. Future research should test this hypothesized effect and extend our findings to other areas exhibiting the social desirability bias.

Overall, construal level theory presents a valuable framework for understanding existing projective measurement techniques. This research paves the way for future research to explore further the link between construal level theory and projective measurement theory. We hope

46 researchers acknowledge the threat posed by the social desirability bias and the versatility and ease of these debiasing techniques by implementing them in their own research.

REFERENCES

Acquisti, Alessandro, Leslie K. John, and George Loewenstein (2012), "The Impact of Relative

Standards on the Propensity to Disclose," Journal of Marketing Research, 49(1), 160-74.

Agrawal, N. and E.W. Wan (2009), "Regulating Risk or Risking Regulation? Construal Levels and

Depletion Effects in the Processing of Health Messages," Journal of Consumer Research, 36 (3),

448-62.

Alpert, Mark I. (1971), "Identification of Determinant Attributes: A Comparison of Methods," Journal of

Marketing Research, 8 (2), 184-91.

Amit, Elinor, Daniel Algom, and Yaacov Trope (2009), "Distance-Dependent Processing of Pictures and

Words," Journal of Experimental Psychology: General, 138 (3), 400.

Anderson, James C. (1978), "The Validity of Haire's Shopping List Projective Technique," Journal of

Marketing Research, 15 (4), 644-49.

Bagozzi, Richard P. and Peter Schnedlitz (1985), "Social Contingencies in the Attitude Model: A Test of

Certain Interaction Hypotheses," Social Psychology Quarterly, 48 (4), 366-73.

Bar-Anan, Yoav, Nira Liberman, Yaacov Trope, and Daniel Algom (2007), "Automatic Processing of

Psychological Distance: Evidence from a Stroop Task," Journal of Experimental Psychology:

General, 136 (4), 610.

Baron, Reuben M. and David A. Kenny (1986), "The Moderatorâ€―Mediator Variable Distinction in

Social Psychological Research: Conceptual, Strategic, and Statistical Considerations," Journal of

Personality and Social Psychology, 51 (6), 1173.

Baumeister, Roy F. (1982), "A Self-Presentational View of Social Phenomena," Psychological bulletin,

91 (1), 3.

Baumgartner, Hans and Jan-Benedict E.M. Steenkamp (1996), "Exploratory Consumer Buying Behavior:

Conceptualization and Measurement," International Journal of Research in Marketing, 13 (2),

121-37.

--- (2001), "Response Styles in Marketing Research: A Cross-National Investigation," Journal of

Marketing Research, 38 (2), 143-56.

Blair, Ed, Seymour Sudman, Norman M. Bradburn, and Carol Stocking (1977), "How to Ask Questions

About Drinking and Sex: Response Effects in Measuring Consumer Behavior," Journal of

Marketing Research, 14 (3), 316-21.

Blake, Brian F., Jillian Valdiserri, Kimberly A. Neuendorf, and Jacqueline Nemeth (2006), "Validity of

the Sds-17 Measure of Social Desirability in the American Context," Personality and Individual

Differences, 40 (8), 1625-36.

Borkenau, Peter and Fritz Ostendorf (1989), "Descriptive Consistency and Social Desirability in Self and

Peer Reports," European journal of personality, 3 (1), 31-45.

Bradburn, Normand M., Seymour Sudman, and Brian Wansink (2004), Asking Questions: The Definitive

Guide to Questionnaire Design: For Market Research, Political Polls, and Social and Health

Questionnaires, Vol. 40: Jossey-Bass Inc Pub.

Bruner, Gordon C. and Anand Kumar (2007), "Gadget Lovers," Journal of the Academy of Marketing

Science, 35 (3), 329-39.

Buckhout, Robert (1965), "Need for Social Approval and Dyadic Verbal Behavior," Psychological

Reports.

Burton, Brian K. and W. Harvey Hegarty (1999), "Some Determinants of Student Corporate Social

Responsibility Orientation," Business & society, 38 (2), 188.

Bush, Victoria D., Gregory M. Rose, Faye Gilbert, and Thomas N. Ingram (2001), "Managing Culturally

Diverse Buyer-Seller Relationships: The Role of Intercultural Disposition and Adaptive Selling in

Developing Intercultural Communication Competence," Journal of the Academy of Marketing

Science, 29 (4), 391-404.

Calder, Bobby J. and Robert E. Burnkrant (1977), "Interpersonal Influence on Consumer Behavior: An

Attribution Theory Approach," The Journal of Consumer Research, 4 (1), 29-38.

Carlson, Les and Sanford Grossbart (1988), "Parental Style and Consumer Socialization of Children,"

Journal of Consumer Research, 15 (1), 77-94.

Catton Jr, W.R. and R.E. Dunlap (1978), "Environmental Sociology: A New Paradigm," The American

Sociologist, 41-49.

Chandran, S and G Menon (2004), "When a Day Means More Than a Year: Effects of Temporal Framing

on Judgments of Health Risk," Journal of Consumer Research, 31 (2), 375-89.

Couzin, Jennifer (2008), "Do Voter Surveys Underestimate the Impact of Racial Bias?," Science, 322

(5899), 180-81.

Crowne, Douglas P. and David Marlowe (1960), "A New Scale of Social Desirability Independent of

Psychopathology," Journal of consulting psychology, 24 (4), 349.

Dawes, Robyn M. and Tom L. Smith (1985), "Attitude and Opinion Measurement," The handbook of

social psychology, 1, 509-66.

Donavan, D. Todd, Tom .J. Brown, and John C. Mowen (2004), "Internal Benefits of Service-Worker

Customer Orientation: Job Satisfaction, Commitment, and Organizational Citizenship Behaviors,"

Journal of Marketing, 68 (1), 128-46.

Dorsch, Michael J., Scott R. Swanson, and Scott W. Kelley (1998), "The Role of Relationship Quality in

the Stratification of Vendors as Perceived by Customers," Journal of the Academy of Marketing

Science, 26 (2), 128.

Dunlop, Riley E. and Kent. D. Van Liere (1978), "The New Environmental Paradigm," Journal of

Environmental Education, 9 (4), 10-19.

Edwards, Allen L. (1957), "The Social Desirability Variable in Personality Assessment and Research."

Fisher, RJ (1993), "Social Desirability Bias and the Validity of Indirect Questioning," Journal of

Consumer Research, 20 (2), 303-15.

Freitas, AL, P Gollwitzer, and Y Trope (2004), "The Influence of Abstract and Concrete Mindsets on

Anticipating and Guiding Others' Self-Regulatory Efforts* 1," Journal of Experimental Social

Psychology, 40 (6), 739-52.

Fujita, Kentaro, Marlone D. Henderson, Juliana Eng, Yaacov Trope, and Nira Liberman (2006), "Spatial

Distance and Mental Construal of Social Events," Psychological Science, 17 (4), 278.

Ganster, Daniel C., Harry W. Hennessey, and Fred Luthans (1983), "Social Desirability Response Effects:

Three Alternative Models," The Academy of Management Journal, 26 (2), 321-31.

Goldsmith, Ronald E. (1987), "Self-Monitoring and Innovativeness," Psychological Reports.

Grubb, EL and BL Stern (1971), "Self-Concept and Significant Others," Journal of Marketing Research,

8 (3), 382-85.

Haire, M (1950), "Projective Techniques in Marketing Research," The Journal of Marketing, 14 (5), 649-

56.

Hegarty, WH and HP Sims (1978), "Some Determinants of Unethical Decision Behavior: An

Experiment," Journal of Applied Psychology, 63 (4), 451-57.

Jayanti, Rama K. and Alvin C. Burns (1998), "The Antecedents of Preventive Health Care Behavior: An

Empirical Study," Journal of the Academy of Marketing Science, 26 (1), 6-15.

John, Leslie K., Alessandro Acquisti, and George Loewenstein (2011), "Strangers on a Plane: Context-

Dependent Willingness to Divulge Sensitive Information," The Journal of Consumer Research,

37 (5), 858-73.

Jones, E.E. and R.E. Nisbett (1972), 5-the Actor and the Observer: Divergent Perceptions of the Causes

of Behavior: General Learning Press.

Josiassen, Alexander (2011), "Consumer Disidentification and Its Effects on Domestic Product Purchases:

An Empirical Investigation in the Netherlands," Journal of Marketing, 75 (2), 124-40.

Khan, Uzma, Meng Zhu, and Ajay Kalra (2011), "When Trade-Offs Matter: The Effect of Choice

Construal on Context Effects," Journal of Marketing Research, 48 (1), 62-71.

Kidwell, Blair, David M. Hardesty, and Terry L. Childers (2008), "Consumer Emotional Intelligence:

Conceptualization, Measurement, and the Prediction of Consumer Decision Making," Journal of

Consumer Research, 35 (1), 154-66.

Kim, Kyeongheui, Meng Zhang, and Xiuping Li (2008), "Effects of Temporal and Social Distance on

Consumer Evaluations," Journal of Consumer Research: An Interdisciplinary Quarterly, 35 (4),

706-13.

King, Maryon F. and Gordon C. Bruner (2000), "Social Desirability Bias: A Neglected Aspect of Validity

Testing," Psychology and Marketing, 17 (2), 79-103.

Kogan, Nathan and Michael A. Wallach (1964), Risk Taking: A Study in Cognition and Personality,

Oxford, England: Holt, Rinehart & Winston.

Kurt, Didem, J.Jeffrey Inman, and Jennifer J. Argo (2011), "The Influence of Friends on Consumer

Spending: The Role of Agency-Communion Orientation and Self-Monitoring," Journal of

Marketing Research, 48 (4), 741-54.

Larsen, Knud S., Harry J. Martin, Richard H. Ettinger, and Joan Nelson (1976), "Approval Seeking,

Social Cost, and Aggression: A Scale and Some Dynamics," The Journal of Psychology, 94 (1),

3-11.

Lastovicka, John L., Lance A. Bettencourt, Renée S. Hughner, and Ronald J. Kuntze (1999), "Lifestyle of

the Tight and Frugal: Theory and Measurement," Journal of Consumer Research, 85-98.

Lenski, Gerhard E. and John C. Leggett (1960), "Caste, Class, and Deference in the Research Interview,"

The American Journal of Sociology, 65 (5), 463-67.

Liberman, N and Y Trope (1998), "The Role of Feasibility and Desirability Considerations in near and

Distant Future Decisions: A Test of Temporal Construal Theory," Journal of Personality and

Social Psychology, 75, 5-18.

Liberman, Nira, Yaacov Trope, and Cheryl J. Wakslak (2007), "Construal Level Theory and Consumer

Behavior," Journal of Consumer Psychology, 17 (2), 113.

Luchs, MG, RW Naylor, JR Irwin, and R Raghunathan (2010), "The Sustainability Liability: Potential

Negative Effects of Ethicality on Product Preference," Journal of Marketing, 74 (5), 18-31.

Luo, Xueming (2007), "Consumer Negative Voice and Firm-Idiosyncratic Stock Returns," Journal of

Marketing, 71 (3), 75-88.

Lusk, Jayson L., Leatta McLaughlin, and Sara R. Jaeger (2007), "Strategy and Response to Purchase

Intention Questions," Marketing Letters, 18 (1), 31-44.

Lynch Jr, John G., Richard G. Netemeyer, Stephen A. Spiller, and Alessandra Zammit (2009), "A

Generalizable Scale of Propensity to Plan: The Long and the Short of Planning for Time and for

Money," Journal of Consumer Research, 37 (1), 108–28.

Maccoby, Eleanor E. and Nathan Maccoby (1954), "The Interview: A Tool of Social Science," Handbook

of social psychology, 1, 449-87.

MacKinnon, David P., Stefany Coxe, and Amanda N. Baraldi (2012), "Guidelines for the Investigation of

Mediating Variables in Business Research," Journal of Business and Psychology, 27 (1), 1-14.

Martin, Wendy K., Seungoog Weun, and Sharon E. Beatty (1994), "Validation of an Impulse Buying

Tendency Scale," in The 1993 annual conference of the Association for Consumer Research,

Nashville, TN.

McFarland, L.A. and A.M. Ryan (2006), "Toward an Integrated Model of Applicant Faking Behavior1,"

Journal of Applied Social Psychology, 36 (4), 979-1016.

Mick, DG (1996), "Are Studies of Dark Side Variables Confounded by Socially Desirable Responding?

The Case of Materialism," Journal of Consumer Research, 23 (2), 106-19.

Mishra, Arul and Himanshu Mishra (2011), "The Influence of Price Discount Versus Bonus Pack on the

Preference for Virtue and Vice Foods," Journal of Marketing Research, 48 (1), 196-206.

Moorman, Christine and Erika Matulich (1993), "A Model of Consumers' Preventive Health Behaviors:

The Role of Health Motivation and Health Ability," The Journal of Consumer Research, 20 (2),

208-28.

Morrison, Donald G. (1979), "Purchase Intentions and Purchase Behavior," The Journal of Marketing,

65-74.

Nancarrow, Clive, Ian Brace, and Len Tiu Wright (2001), "" Tell Me Lies, Tell Me Sweet Little Lies":

Dealing with Socially Desirable Responses in Market Research," The Marketing Review, 2 (1),

55-69.

Naylor, Rebecca Walker, Cait Poynor Lamberton, and David A. Norton (2011), "Seeing Ourselves in

Others: Reviewer Ambiguity, Egocentric Anchoring, and Persuasion," Journal of Marketing

Research, 48 (3), 617-31.

Nigro, G and U Neisser (1983), "Point of View in Personal Memories* 1," Cognitive psychology, 15 (4),

467-82.

Nussbaum, Shiri, Yaacov Trope, and Nira Liberman (2003), "Creeping Dispositionism: The Temporal

Dynamics of Behavior Prediction," Journal of Personality and Social Psychology, 84 (3), 485.

Paulhus, Delroy L. (1991), "Measurement and Control of Response Bias," in Measures of Personality and

Social Psychological Attitudes, Measures of Social Psychological Attitudes, Vol. 1, ed. John P.

Robinson, Phillip R. Shaver and Lawrence S. Wrightsman, San Diego, CA: Academic Press, 17-

59.

Paulhus, Delroy L. (2002), "Socially Desirable Responding: The Evolution of a Construct," The role of

constructs in psychological and educational measurement, 49–69.

Paulhus, Delroy L. and M. Nadine Bruce (1992), "The Effect of Acquaintanceship on the Validity of

Personality Impressions: A Longitudinal Study," Journal of Personality and Social Psychology,

63 (5), 816.

Paulhus, Delroy L. and Shawn Reynolds (1995), "Enhancing Target Variance in Personality Impressions:

Highlighting the Person in Person Perception," Journal of Personality and Social Psychology, 69

(6), 1233.

Pauls, Cornelia A. and Nicolas W. Crost (2004), "Effects of Faking on Self-Deception and Impression

Management Scales," Personality and Individual Differences, 37 (6), 1137-51.

Pham, Michel Tuan (1998), "Representativeness, Relevance, and the Use of Feelings in Decision

Making," Journal of Consumer Research, 144-59.

Phillips, D.L. and K.J. Clancy (1972), "Some Effects Of" Social Desirability" In Survey Studies," The

American Journal of Sociology, 77 (5), 921-40.

Preacher, Kristopher J. and Andrew F. Hayes (2008), "Asymptotic and Resampling Strategies for

Assessing and Comparing Indirect Effects in Multiple Mediator Models," Behavior research

methods, 40 (3), 879-91.

Pronin, Emily and Lee Ross (2006), "Temporal Differences in Trait Self-Ascription: When the Self Is

Seen as an Other," Journal of Personality and Social Psychology, 90 (2), 197.

Regan, Pamela C., Mark Snyder, and Saul M. Kassin (1995), "Unrealistic Optimism: Self-Enhancement

or Person Positivity?," Personality and Social Psychology Bulletin, 21 (10), 1073.

Richins, M.L. (2004), "The Material Values Scale: Measurement Properties and Development of a Short

Form," Journal of Consumer Research, 209-19.

Richins, ML (1983), "An Analysis of Consumer Interaction Styles in the Marketplace," Journal of

Consumer Research, 10 (1), 73-82.

Rim, SoYon, James S. Uleman, and Yaacov Trope (2009), "Spontaneous Trait Inference and Construal

Level Theory: Psychological Distance Increases Nonconscious Trait Thinking," Journal of

Experimental Social Psychology, 45 (5), 1088-97.

Rindfleisch, Aric, Alan J. Malter, Shankar Ganesan, and Christine Moorman (2008), "Cross-Sectional

Versus Longitudinal Survey Research: Concepts, Findings, and Guidelines," Journal of

Marketing Research, 45 (3), 261-79.

Rindfleisch, Aric, Nancy Y. Wong, and James E. Burroughs (2006), "Seeking Certainty Via Brands: An

Examination of Materialism and Brand Resonance," forthcoming in Advances in Consumer

Research, 34.

Robertson, Dan H. and Robert W. Joselyn (1974), "Projective Techniques in Research," Journal of

Advertising Research.

Rook, Dennis W. (1987), "The Buying Impulse," The Journal of Consumer Research, 14 (2), 189-99.

Roth, Marcus and Philipp Y. Herzberg (2007), "The Resilient Type:Simply the Best'or Merely an Artifact

of Social Desirability?," Psychology Science, 49 (2), 150.

Sabourin, Sabourin, Lise Bourgeois, Pierre Gendreau, and Monique Morval (1989), "Self-Deception,

Impression Management, and Consumer Satisfaction with Mental Health Treatment,"

Psychological Assessment: A Journal of Consulting and Clinical Psychology, 1 (2), 126.

Sackeim, Harold A. and Ruben C. Gur (1979), "Self-Deception, Other-Deception, and Self-Reported

Psychopathology," Journal of Consulting and Clinical Psychology, 47 (1), 213.

Schuman, Howard and Stanley Presser (1981), Questions and Answers in Attitude Surveys: Experiments

on Question Form, Wording, and Context: Sage Publications, Inc.

Sen, Sankar and C.B. Bhattacharya (2001), "Does Doing Good Always Lead to Doing Better? Consumer

Reactions to Corporate Social Responsibility," Journal of Marketing Research, 38 (2), 225-43.

Steenkamp, Jan-Benedict E.M., Martijn G. de Jong, and Hans. Baumgartner (2010), "Socially Desirable

Response Tendencies in Survey Research," Journal of Marketing Research, 47 (2), 199-214.

Steenkamp, Jan-Benedict E.M., Frenkel ter Hofstede, and Michel Wedel (1999), "A Cross-National

Investigation into the Individual and National Cultural Antecedents of Consumer

Innovativeness," The Journal of Marketing, 55-69.

Stephan, Elena, Nira Liberman, and Yaacov Trope (2010), "The Effects of Time Perspective and Level of

Construal on Social Distance," Journal of Experimental Social Psychology, 47 (2), 397-402.

Strahan, Robert and Kathleen C. Gerbasi (1973), "Semantic Style Variance in Personality

Questionnaires," The Journal of Psychology, 85 (1), 109-18.

Sudman, Seymour, Norman M. Bradburn, and Norbert Schwarz (1996), Thinking About Answers: The

Application of Cognitive Processes to Survey Methodology: Jossey-Bass.

Tian, Kellly Tepper, William O. Bearden, and Garry L. Hunter (2001), "Consumers' Need for

Uniqueness: Scale Development and Validation," Journal of Consumer Research, 50-66.

Tourangeau, Roger and Ting Yan (2007), "Sensitive Questions in Surveys," Psychological bulletin, 133

(5), 859.

Trope, Y and N Liberman (2003), "Temporal Construal," Psychological Review, 110 (3), 403-20.

Trope, Y, N Liberman, and C Wakslak (2007), "Construal Levels and Psychological Distance: Effects on

Representation, Prediction, Evaluation, and Behavior," Journal of Consumer Psychology, 17 (2),

83–95.

Trope, Yaacov and Nira Liberman (2010), "Construal-Level Theory of Psychological Distance,"

Psychological Review, 117 (2), 440.

Trudel, Remi and June Cotte (2008), "Does Being Ethical Pay?," Wall Street Journal, 1.

Unger, Lynette S. and Jerome B. Kernan (1983), "On the Meaning of Leisure: An Investigation of Some

Determinants of the Subjective Experience," The Journal of Consumer Research, 9 (4), 381-92.

Wakslak, C and Y Trope (2009), "The Effect of Construal Level on Subjective Probability Estimates,"

Psychological Science, 20 (1), 52.

Wakslak, Cheryl J., Shiri Nussbaum, Nira Liberman, and Yaacov Trope (2008), "Representations of the

Self in the near and Distant Future," Journal of Personality and Social Psychology, 95 (4), 757.

Weaver, Kimberlee and Norbert Schwarz (2008), "Self-Reports in Consumer Research," Handbook of

Consumer Psychology, 1081-102.

Weinstein, Neil D. (1980), "Unrealistic Optimism About Future Life Events," Journal of Personality and

Social Psychology, 39 (5), 806.

White, Katherine, Rhiannon MacDonnell, and Darren W. Dahl (2011), "It's the Mind-Set That Matters:

The Role of Construal Level and Message Framing in Influencing Consumer Efficacy and

Conservation Behaviors," Journal of Marketing Research, 48 (3), 472-85.

Wiggins, Jerry S. (1959), "Interrelationships among Mmpi Measures of Dissimulation under Standard and

Social Desirability Instruction," Journal of consulting psychology, 23 (5), 419.

Wright, S., C. Manolis, D. Brown, X. Guo, J. Dinsmore, C.Y.P. Chiu, and F.R. Kardes (2011),

"Construal-Level Mind-Sets and the Perceived Validity of Marketing Claims," Marketing Letters,

1-9.

Yan, D. and J. Sengupta (2011), "Effects of Construal Level on the Price-Quality Relationship," Journal

of Consumer Research, 38 (2), 376-89.

Zerbe, Wilfred J. and Delroy L. Paulhus (1987), "Socially Desirable Responding in Organizational

Behavior: A Reconception," The Academy of Management Review, 12 (2), 250-64.

Zhao, Xinshu, John G. Lynch, and Qimei Chen (2010), "Reconsidering Baron and Kenny: Myths and

Truths About Mediation Analysis," Journal of Consumer Research, 37 (2), 197-206.

TABLES AND FIGURES

TABLE 1

SUMMARY OF ARTICLES DOCUMENTING THE SOCIAL DESIRABILITY BIAS (JCR, JMR, JM, JAMS: 1998-2011)

Social Correlation Desirability Coefficients Type of Use Bias (Content/SDB Source of Content Author(s) Source Reported Scale Scales) Scales Interpretation (John, Acquisti, JCR (2011) Situational One item Not reported. Not reported. The authors pretest the and Analysis scale asking perceived social Loewenstein) if behavior desirability of various is ‗cool‘ or behavioral measures not. before examining when respondents divulge socially sensitive information. (Lynch Jr et al.) JCR (2009) Scale MCSDS = -.05 to .41 New scales; each The long-run money constructi measuring version of the planning on propensity to scale suggests social plan desirability bias. The remaining scales did not suggest socially desirable responding. (Rindfleisch et JCR (2008) Scale BIRD =-.39 to .22 Several existing Measures of materialism, al.) constructi multi-item scales personal insecurity, social on insecurity, developmental insecurity, existential insecurity, age, gender, and brand loyalty all exhibit the social desirability bias. (Kidwell, JCR (2008) Scale MCSDS =.05 New Consumer The authors conclude that Hardesty, constructi Emotional the CEIS does not appear and Childers) on Intelligence to be susceptible to the Scale (CEIS) social desirability bias. (Richins) JCR (2004) Scale MCSDS =-.16 to .06 Short form of the The short form of the MV is developme Material Values not influenced by the nt (MV) Scale social desirability bias. (Tian, Bearden, JCR (2001) Scale MCSDS & =.1 (MCSDS) Consumer‘s need Results indicate that the and Hunter) developme The IM for uniqueness social desirability bias is nt factor of the =.01 (IM) scale not a threat to the BIDR measure‘s validity. Version 6 (Lastovicka et JCR (1999) Scale MCSDS & =.17 (MCSDS) Frugality scale. The frugality measure is al.) constructi BIRD independent from on =-.14 (BIRD) response bias. However, measures of compulsive buying and price conscious exhibit the social desirability bias. The authors report a significant negative correlation (r = -.33) between the social desirability bias and acquiescence bias. (Pham) JCR (1998) Behavioral Not reported Not reported New multi-item Although the author used bias scales measuring indirect questioning to behavioral reduce the social

intentions and desirability bias, he perceived incorporated a measure of motives. the social desirability bias as a covariate in Study 1. (Mishra and JMR (2011) Situational MCSDS Not reported. A discrete choice The authors did not find any Mishra) Analysis between product influence of socially promotions desirable responding (price discount across the experimental versus bonus conditions. packs versus indifferent). (Luo) JM (2007) Situational Strahan & Not reported. Measures of The authors conducted a analysis Gerbasi‘s competitor series of comparative (1973) orientation, tests indicating that the shortened competitor influence of competitor version of alliances, and orientation and CA on the MCSDS firm ROE. firm ROE is similar in models that include the social desirability bias versus models that do not include this variable. Thus, they conclude that the social desirability bias does not seem to be a major concern. (Josiassen) JM (2011) Scale MCSDS =-.02 New consumer The authors concluded that constructi disidentification a non-significant on scale (CDI) correlation between CDI and the social desirability bias scale suggests that social desirability bias is not a concern. (Donavan, JM (2004) Situational Strahan & =.09 to .3 Several existing The authors concluded that Brown, and analysis Gerbasi‘s multi-item scales the social desirability bias Mowen) (1973) including job had a significant effect on shortened satisfaction, customer orientation and version of commitment, commitment. However, the MCSDS customer the bias did little to orientation, and change the structural organization paths in their SEM citizenship model. behaviors. (Steenkamp, JM (1999) Situational MCSDS =.01 The consumer- The authors concluded that Hofstede, analysis specific a non-significant and Wedel) Exploratory correlation between EAP Acquisition of and the social desirability Products (EAP) bias scale suggests that scale the social desirability bias (Baumgartner is not a concern. and Steenkamp 1996). (Bruner and JAMS (2007) Scale MCSDS =.02 A new trait scale The authors concluded that Kumar) constructi identifying a non-significant on individuals with correlation between their an intrinsic gadget lover measure and motivation to the social desirability bias adopt novel tech scale suggests that the products (gadget social desirability bias is lovers). not a concern. (Bush et al.) JAMS (2001) Situational Strahan & Not reported. Several existing The authors report non- analysis Gerbasi‘s multi-item scales significant correlations (at (1973) including the .05 level) between the

shortened empathy, constructs of interest and version of worldmindednes the social desirability bias the MCSDS s, ethnocentrism, scale. attributional complexity, adaptive selling, and perceived intercultural communication competence. (Dorsch, JAMS (1998) Situational Carlson & Not reported. Buyer-vendor The authors found that their Swanson, analysis Grossbart‘s perceptions social desirability bias and Kelley) (1988) scale did not account for shortened any appreciable version of variability in buyer- the MCSDS vendor perceptions, Λ =.88, F(12, 79) = .88, p = .57. Thus, the authors conclude that these findings are not likely to be significantly influenced by socially desirable responding.

TABLE 2

DESCRIPTIVE STATISTICS (STUDY 1)

% or mean SD Male 55% - Country: India 53% - United States 19% - Canada 1% - UK and Northern Ireland 1% - Philippines 1% - Other 6% - 19% - Completion Rate 83% - Completion Time 9 min., 43 sec. 8 min., 55 sec. Income: Below $20,000 43% - $20,000 - $29,999 15% - $30,000 - $39,999 9% - $40,000 - $49,999 5% - $50,000 - $59,999 4% - $60,000 - $69,999 3% - $70,000 - $79,999 3% - $80,000 - $89,999 1% - $90,000 or more 2% - 15% - Education: Less than High School 1% - High School / GED 6% - Some College 15% - 2-year College Degree 11% - 4-year College Degree 29% - Masters Degree 20% - Doctoral Degree 1% - Professional Degree (JD, MD) 2% - 17% SDE 4.54 1.04 IM 4.57 1.06 Materialism 4.34 0.98 Frequency of Healthy Preventative Behaviors 4.99 1.12 Impulsive Buying Tendency 4.12 0.96 Corporate Social Responsibility (CSR) 4.17 1.01 Green Orientation 4.68 0.84 Age (measured via DOB) 29.89 9.44 Television Viewing Frequency 5.03 1.29 Regulation of Alcohol Consumption 4.66 2.0

Note: With listwise deletion of missing or incomplete cases, N = 453

TABLE 3

SUMMARY OF STANDIZED TOTAL, DIRECT, AND INDIRECT EFFECTS FOR EACH PRODUCT CATEGORY (STUDY 3)

Complete sample (N = 427)

Coefficient (SE) Lower CI Upper CI Direct effect of interaction on SDE (IV  M1) .02 (.1) Direct effect of interaction on IM (IV  M2) .27** (.1)

Choice Between Laundary Detergents (DV1) Direct effect of SDE on choice (M1  DV1) -.13 (.11) Direct effect of IM on choice (M2  DV1) .28* (.11) Total effect of interaction on choice (IV  DV1) -.64 (.41) Total indirect effect .07* (.04) 0.010 0.195 Indirect effect through SDE (IV  M1  DV1) 0 (.02) -0.053 0.025 Indirect effect through IM (IV  M2  DV1) .08** (.05) 0.013 0.204

Choice Between Hand Sanitizers (DV2) Direct effect of SDE on choice (M1  DV2) -.22 (.13) Direct effect of IM on choice (M2  DV2) .28* (.13) Total effect of interaction on choice (IV  DV2) -.22 (.23) Total indirect effect .07* (.05) 0.001 0.188 Indirect effect through SDE (IV  M1  DV3) -.06 (.04) -0.065 0.040 Indirect effect through IM (IV  M2  DV3) .08* (.05) 0.009 0.206

*Significant at 5 % level. **Significant at 1 % level..

Significance levels are based on BC bootstraps (n = 5,000) 95% CIs using Preacher and Hayes's (2008) Indirect Macro.

FIGURE 1

STUDY 1: PLOTS OF WITHIN EXPERIMENTAL CONDITION STANDARDIZED REGRESSION COEFFICIENTS OF SDE AND IM FOR SEVEN MARKETING SCALES

FIGURE 1 CONTINUED

APPENDIX A Manipulations and Measures of Study 1

ANONYMOUS CONDITION (Public or Control Condition)

Introduction

Thank you for participating in this research study.

Please note that the responses you provide in this survey will be kept completely confidential.

In order to proceed with the survey you must acknowledge that you've read and understand the above statement. Do you understand that your responses will be kept fully confidential (if you do not wish to proceed please select 'No')?

PUBLIC CONDITION (Anonymous or Control Condition)

Introduction

Thank you for participating in this research study.

Please note that the responses you provide in this survey will be posted online along with personal identifying information such as your name, which is validated by your IP address.

In order to proceed with the survey you must acknowledge that you've read and understand the above statement. Do you understand that your responses will be made publicly available (if you do not wish to proceed please select 'No')?

CONTROL CONDITION (Anonymous or Non-Anonymous)

Introduction

Thank you for participating in this research study. Please begin by clicking on the "NEXT" button below.

DIRECT MEASURES (Indirect Using Temporal Distance or Indirect Using Spatial Distance or Indirect Using Social Distance)

MATERIALISM (Richins 2004)

Please indicate how much you agree or disagree with the following statements.

FREQUENCY OF HEALTHY PREVENTATIVE BEHAVIORS (Jayanti and Burns 1998)

Please indicate how often you engage in the following activities.

TELEVISION VIEWING FREQUENCY

Please indicate how often you engage in the following activity.

REGULATION OF ALCOHOL CONSUMPTION (Jayanti and Burns 1998)

Please indicate how often you engage in the following activity.

IMPULSIVE BUYING TENDENCY (Martin et al. 1994)

Please indicate the degree to which the following statements describe your behavior when you are shopping at a mall. Think about the last few times you have been shipping in a mall. Please indicate how much you agree or disagree with each statement.

CORPORATE SOCIAL RESPONSIBILITY (Sen and Bhattacharya 2001)

Please indicate how much you agree or disagree with the following statements.

GREEN ORIENTATION (Catton Jr and Dunlap 1978)

Please indicate how much you agree or disagree with the following statements.

INDIRECT MEASURES USING TEMPORAL DISTANCE (Indirect Using Spatial Distance or Indirect Using Social Distance or Direct)

MATERIALISM

A year from now, how much will you agree or disagree with the following statements. How would you respond a year from now?

FREQUENCY OF HEALTHY PREVENTATIVE BEHAVIORS

A year from now, how often are you likely to engage in the following activities.

TELEVISION VIEWING FREQUENCY

A year from now, how often are you likely to engage in the following activity.

REGULATION OF ALCOHOL CONSUMPTION

A year from now, how often are you likely to engage in the following activity.

IMPULSIVE BUYING TENDENCY

Please indicate the degree to which the following statements would describe your behavior assuming you are shopping at a mall a year from now. With this in mind, please indicate how much you agree or disagree with each statement.

CORPORATE SOCIAL RESPONSIBILITY

A year from now, how much will you agree or disagree with the following statements. How would you respond a year from now?

GREEN ORIENTATION

A year from now, how much will you agree or disagree with the following statements. How would you respond a year from now?

INDIRECT MEASURES USING SPATIAL DISTANCE (Indirect Using Temporal Distance or Indirect Using Social Distance or Direct)

MATERIALISM

Let us assume you are in another area, far from where you are now. With this in mind, please indicate how much you agree or disagree with the following statements.

FREQUENCY OF HEALTHY PREVENTATIVE BEHAVIORS

Let us assume you are in another area, far from where you are now. With this in mind, how often would you engage in the following activities?

TELEVISION VIEWING FREQUENCY

Let us assume you are in another area, far from where you are now. With this in mind, how often would you engage in the following activity?

REGULATION OF ALCOHOL CONSUMPTION

Let us assume you are in another area, far from where you are now. With this in mind, how often would you engage in the following activity?

IMPULSIVE BUYING TENDENCY

Please indicate the degree to which the following statements describe your behavior assuming you are shopping at a mall far from where you are now. With this in mind, please indicate how much you agree or disagree with each statement.

CORPORATE SOCIAL RESPONSIBILITY

Let us assume you are in another area, far from where you are now. With this in mind, please indicate how much you agree or disagree with the following statements.

GREEN ORIENTATION

Let us assume you are in another area, far from where you are now. With this in mind, please indicate how much you agree or disagree with the following statements.

INDIRECT MEASURES USING SOCIAL DISTANCE (Indirect Using Temporal Distance or Indirect Using Spatial Distance or Direct)

MATERIALISM

Please indicate how much you agree or disagree with the following statements.

FREQUENCY OF HEALTHY PREVENTATIVE BEHAVIORS

Please indicate how often most people engage in the following activities.

TELEVISION VIEWING FREQUENCY

Please indicate how often most people engage in the following activity.

REGULATION OF ALCOHOL CONSUMPTION

Please indicate how often most people engage in the following activity.

IMPULSIVE BUYING TENDENCY

Please indicate the degree to which the following statements describe most people's behavior when shopping at a mall. Think about the last few times you have been shopping in a mall. Please indicate how much you agree or disagree with each statement.

CORPORATE SOCIAL RESPONSIBILITY

Please indicate how much most people would agree or disagree with the following statements.

GREEN ORIENTATION

Please indicate how much most people would agree or disagree with the following statements.

PERCEIVED ANONYMITY SCALE

Please indicate how much you agree or disagree with the following statements.

DEMOGRAPHIC MEASURES

Please answer the following questions about yourself.

What is your gender?

What year were you born?

Is English your native language?

What is the highest level of education you have completed?

What is your annual income range?

In which country do you reside?

APPENDIX B Manipulations and Measures of Study 2

CONSTRUAL MIND-SET MANIPULATION (Concrete or Abstract) (Freitas et al. 2004)

CHOICE TASK (ADAPTED FROM LUCHS ET AL., 2010)

PRODUCT EVALUTIONS (ADAPTED FROM LUCHS ET AL., 2010)

DEMOGRAPHIC MEASURES

Please answer the following questions about yourself.

What is your gender?

What year were you born?

APPENDIX C Manipulations and Measures of Study 3

SELF-PRESENTATION CONCERN (Present or Control)

PRODUCT OPTIONS (ADAPTED FROM LUCHS ET AL., 2010) Randomized Presentation

100

101

CHOICE TASK (Psychologically Proximal or Psychologically Distant) Randomized Brand Presentation

\ Seventh Generation

\ Tide

\ Seventh Generation \ Tide

102

DEMOGRAPHIC MEASURES

Please answer the following questions about yourself.

What is your gender?

What year were you born?

Is English your native language?

What is the highest level of education you have completed?

What is your annual income range?

103

104

APPENDIX D Individual Difference Measures

105

THE BALANCED INVENTORY OF DESIRABLE RESPONDING (BIRD) SCALE (Paulhus 1991) (Used in Study 1 & Study 3)

Please indicate how much you agree or disagree with the following statements.

106

IMPRESSION MANAGEMENT (ADAPTED FROM LUCHS ET AL., 2010) (Used in Study 2 & Study 3)

107