Understanding the Effects of Social Norms and Knowledge on Socially Responsible Consumer Behavior (SRCB)

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Tae-Im Han

Graduate Program in Fashion and Retail Studies

The Ohio State University

2014

Dissertation Committee:

Professor Leslie Stoel, Advisor

Associate Professor Catherine Montalto

Associate Professor Sharon Seiling

Copyrighted by

Tae-Im Han

2014

Abstract

Growing consumer awareness about the environmental and social impact of consumption has led to an increase in demand for more socially responsible product alternatives. This dissertation examined diverse aspects of consumer behaviors associated with socially responsible products to contribute to a better understanding of socially responsible consumer behavior (SRCB). It consisted of three main studies and a major focus was placed on examining whether, and to what extent, social norms and product knowledge affect purchase of socially responsible products. Study 1 conducted a meta- analysis of previous consumer studies involving a wide range of product types and ethical issues. Study 2 examined the influence of different types of social norms and product knowledge on consumers’ purchase of organic cotton and fair-trade apparel. Structural equation modeling (SEM) and ANOVA were used to test the cause-effect relationships among constructs and to compare the results between the two product groups. Study 3 examined the efficacy of using social media to increase product knowledge and promote

SRCB, particularly focusing on the role of information format and information characteristics. Similar to Study 2, SEM and ANOVA were used to test the hypotheses of

Study 3.

All three studies from this dissertation shared a topic of identifying significant factors that influence purchase decisions of socially responsible products. Each study is ii designed to build upon and complement the findings from other studies. To begin with,

Study 1 reviewed applications of theory of planned behavior (TPB) in previous SRCB literature and verified the theory’s efficacy to explain purchase behaviors in this study domain. An interesting finding from Study 1 was that consumers’ purchase intention had an especially strong relationship with subjective norms which was in contrast to previous studies that suggested a weak association between subjective norms and behavioral intention. This result indicated that the influence of significant others in one’s social environment is critical in the SRCB context.

Study 2 enhanced our knowledge about the influence of social norms on SRCB by distinguishing the concept into two distinct types: injunctive and descriptive norms.

Injunctive norms were more effective in encouraging positive attitudes while descriptive norms had a stronger effect on increasing purchase intentions. The two types of social norms had independent effects on purchase behavior of socially responsible apparel products. Additionally, Study 2 confirmed the important role of knowledge in the purchase process of these products. Both objective and subjective knowledge significantly influenced attitude toward purchasing the products.

Hence, taking into account such findings that confirm the important role of social norms and product knowledge, Study 3 examined ways to increase knowledge level and to motivate purchase related to socially responsible products, particularly by testing the effectiveness of information formats and information characteristics which are commonly features of social media. The results demonstrated a significant impact of information

iii format on consumers’ cognitive and affective responses and information characteristics

(i.e., information quantity and quality) on knowledge enhancement.

This dissertation as a whole contributed to a better understanding of socially responsible consumers by providing detailed information about the effect of social norms and product knowledge on their purchase decisions. Based on the findings, theoretical and practical implications are presented. Potential limitations of these studies and future research directions are also suggested.

Acknowledgments

I would like to express the deepest appreciation to my advisor, Dr. Leslie Stoel.

Her constant support and encouragement have made my experience in graduate school an enjoyable and rewarding journey. This dissertation would not have been possible without her guidance. I would also like to thank Dr. Catherine Montalto and Dr. Sharon Seiling for serving as my committee members. Their constructive suggestions and comments have helped me develop ideas and expand my understanding of the research topic.

Furthermore, I would like to thank my fellow graduate students from the

Consumer Sciences program. They have been a great resource from the beginning of my

Ph.D. studies. Hwe-Won Kim has also made my life at Columbus more meaningful. I will always fondly remember our stimulating discussions while rushing to meet our deadlines and all the fun we had together in the last four years.

Finally, I must thank my family. Words cannot express how grateful I am to each member of my family for their love, help, and sacrifice throughout my studies. I dedicate this dissertation to my mother, Yehwa Jun, who has always encouraged me to enjoy the process of learning and to follow my dreams wherever they might lead me.

Vita

March, 1980 ...... Born- Seoul, Korea

2005...... B.S. Clothing and Textiles,

Ewha Womans University

2010...... M.A. Clothing and Textiles,

Ewha Womans University

2010 to 2013 ...... Graduate Teaching Associate,

Department of Human Sciences,

The Ohio State University

Publications

Research Publications

1. Han, T. & Rudd, N. (accepted). Images of beauty: Race, gender, age, and occupational analysis of fashion magazine covers. Journal of Global Fashion Marketing. doi: 10.1080/20932685.2014.926131.

2. Han, T. & Chung, J. (2014). Korean consumers' motivations and perceived risks toward the purchase of organic cotton apparel. Clothing and Textiles Research Journal, 32(4), 235-250.

3. Han, T. & Cho, K. (2010). The marketing strategies and design styles of fast fashion companies, The Korean Society of Fashion Business, 10(3), 21-34.

Referred Proceedings vi

1. Han, T. & Stoel, L. (November, 2014). Explaining socially responsible consumer behavior: A meta-analytic review of theory of planned behavior. Proceedings of the International Textiles and Apparel Association (ITAA) Annual Conference, Charlotte, North Carolina.

2. Han, T. & Stoel, L. (November, 2013). A typology of consumers’ familiarity and experience of organic cotton apparel. Proceedings of the International Textiles and Apparel Association (ITAA) Annual Conference, New Orleans, LA.

3. Han, T. & Kandampully, J. (July, 2013). Socially responsible consumer behavior: A literature review of theory of planned behavior. Proceedings of the 20th International EIRASS Conference on Recent Advances in Retailing and Service Sciences, Philadelphia, PA.

4. Han, T. & Kandampully, J. (July, 2013). Motivating consumers’ socially responsible behavior: Efficacy of using normative messages on Facebook. Proceedings of the 20th International EIRASS Conference on Recent Advances in Retailing and Service Sciences, Philadelphia, PA.

5. Han, T. & Rudd, N. (November, 2012). Images of beauty: Race, gender, age, and occupational analysis of fashion magazine covers. Proceedings of the International Textiles and Apparel Association (ITAA) Annual Conference, Honolulu, HI.

6. Han, T. & Chung, J. (April 2012). Role of organic cues on purchase intention. Proceedings of the AMA/ACRA Triennial Retail Conference, Seattle, WA.

Fields of Study

Major Field: Fashion and Retail Studies

vii

Table of Contents

Abstract ...... ii

Acknowledgments...... v

Vita ...... vi

List of Tables ...... xiv

List of Figures ...... xvi

Chapter 1: Introduction ...... 1

1.1. Overview ...... 1

1.2. Problem Statement ...... 2

1.3. Purpose ...... 3

1.4. Definition of Terms ...... 7

Chapter 2: Study 1- Explaining Socially Responsible Consumer Behavior: A Meta-

Analytic Review of Theory of Planned Behavior ...... 12

2.1. Overview ...... 12

2.2. Problem Statement ...... 13

2.3. Purpose of Study ...... 14

viii

2.4. Literature Review ...... 21

2.4.1. Theory of planned behavior ...... 21

2.4.2. Socially Responsible Consumers...... 24

2.5. Methods ...... 27

2.5.1. Sample of Studies ...... 27

2.5.2. Data Coding and Data Retrieval ...... 37

2.5.3. Meta-Analytic Strategy ...... 37

2.6. Results ...... 38

2.7. Discussion ...... 49

Chapter 3: Study 2- The Effect of Social Norms and Product Knowledge on Purchase

Behaviors of Organic Cotton and Fair-Trade Apparel ...... 56

3.1. Overview ...... 56

3.2. Problem Statement ...... 57

3.3. Purpose of Study ...... 59

3.4. Literature Review ...... 59

3.4.1. Theory of Planned Behavior ...... 59

3.4.2. Focus Theory of Normative Conduct ...... 61

3.4.3. Product Knowledge ...... 62

3.5. Conceptual Model and Hypotheses Development ...... 64

3.5.1. Effect of Attitude and Perceived Behavioral Control on Purchase Intention .. 64

3.5.2. Effect of Social Norms on Attitude and Purchase Intention ...... 65

3.5.3. Effect of Product Knowledge on Attitude and Perceived Behavioral Control 66

3.5.4. Effect of Ethical Issue Type on Consumer Response...... 68

3.6. Methods ...... 69

3.6.1. Identification of Socially Responsible Products ...... 69

3.6.2. Instruments ...... 74

3.6.3. Procedure ...... 76

3.6.4. Sample ...... 76

3.6.5. Data Analysis ...... 80

3.7. Results ...... 80

3.7.1. Measurement Model ...... 80

3.7.2. Structural Model ...... 89

3.7.3. Hypotheses Tests ...... 90

3.8. Discussion ...... 98

3.9. Managerial Implications ...... 103

3.10. Theoretical Implications ...... 105

3.11. Limitations and Suggestions for Future Research...... 106

Chapter 4: Study 3- Efficacy of Using Social Media to Increase Knowledge and Promote

SRCB ...... 107

4.1. Overview ...... 107

4.2. Problem Statement ...... 108

4.3. Purpose of Study ...... 109

4.4. Literature Review ...... 110

4.4.1. Media Richness...... 110

4.4.2. Hierarchy of Effects Model ...... 111

4.4.3. Information Quality and Quantity ...... 112

4.4.4. Objective and Subjective Knowledge...... 114

4.5. Conceptual Model and Hypotheses ...... 115

4.5.1. Effect of Media Richness on Consumer Response ...... 115

4.5.2. Effect of Information Quantity and Quality on Knowledge ...... 117

4.5.3. Relationships of Knowledge, Emotion, and Purchase Intention ...... 118

4.6. Methods ...... 120

4.6.1. Instrument ...... 120

4.6.2. Procedures ...... 122

4.6.3. Sample ...... 125

4.6.4. Data Analysis ...... 128

4.7. Results ...... 128

4.7.1. ANOVA ...... 128

4.7.2. Measurement Model ...... 129

4.7.3. Structural Model ...... 132

4.7.4. Hypotheses Tests (H2-H4) ...... 136

4.8. Discussion ...... 141

4.9. Managerial Implications ...... 145

4.10. Theoretical Implications ...... 146

4.11. Limitations and Suggestions for Future Research...... 146

Chapter 5: Discussion ...... 149

5.1. Overview ...... 149

5.2. General Discussion ...... 149

5.3. Theoretical Implications ...... 152

5.4. Practical Implications ...... 156

5.5. Limitations and Suggestions for Future Research...... 158

References ...... 161

Appendix A: Study2 (Preliminary Study) Questionnaire ...... 177

Appendix B: Study2 (Main Study) Questionnaire ...... 181

Appendix C: Study3 Questionnaire ...... 195

xii

Appendix D: Text +Image Posting ...... 206

Appendix E: Text Only Posting ...... 209

Appendix F: Willingness to Pay and Willingness to Search ...... 211

xiii

List of Tables

Table 1. Description of the Sample of Studies and Data Sets ...... 30

Table 2. Descriptive Statistics of the Studies...... 39

Table 3. Meta-analysis of the TPB components ...... 41

Table 4. Hierarchical Regression of Intention on TPB and Moral Norms ...... 44

Table 5. Hierarchical Regression of Intention on TPB and Self-Identity ...... 44

Table 6. Hierarchical Regression of Intention on TPB and Environmental Consciousness

...... 44

Table 7. Meta-Analysis of Moderators ...... 48

Table 8. Results of Between-Group Homogeneity Statistic ...... 49

Table 9. Domains of Scales that Measure Individual’s Level of Socially Responsible

Consumption ...... 71

Table 10. Demographic Summary Statistics of the Sample (n=500)...... 79

Table 11. Results of the Final Measurement Models for the Two Groups ...... 82

Table 12. AVEs, Correlations, and Squared Correlations of the Measurement Models .. 87

Table 13. Comparison of Path Coefficients ...... 94

Table 14. Ethical Issue Main Effects ...... 95

Table 15. Ethical Issue Moderating Effects ...... 96

Table 16. Demographic Summary Statistics of the Sample (n=455)...... 127 xiv

Table 17. Media Richness Main Effects (ANOVA) ...... 129

Table 18. Result of Measurement Model ...... 133

Table 19. Measurement model: Correlations between Latent Variables ...... 136

Table 20. Hypotheses Testing Results ...... 139

Table 21. Standardized Total Effects, Direct Effects, and Indirect Effects ...... 140

List of Figures

Figure 1. Conceptual Model of Study 2 ...... 64

Figure 2. Confirmatory Factor Analysis Model of Study 2 ...... 88

Figure 3. Summary of SEM Results of Study 2...... 97

Figure 4. Conceptual Model of Study 3 (Hierarchy of Effects Model) ...... 115

Figure 5. Confirmatory Factor Analysis Model of Study 3 ...... 131

Figure 6. Summary of SEM Results of Study 3...... 138

xvi

Chapter 1: Introduction

1.1. Overview

Consumers are concerned about the impact their consumption has on the social and natural environment. A recent study by Boston Consulting Group (BCG) querying

9,000 adults from nine countries showed that consumers are concerned about ethical issues and believe it is important for companies to be socially responsible (Manget,

Roche, & Munnich, 2009). Accordingly, consumer interest in products with certifications such as “Eco-friendly”, “Fair-trade”, or “Child Labor Free” is on the rise (Carrigan & De

Pelsmacker, 2009) and being translated into purchasing decisions in the current market.

There's no doubt that many consumers worldwide have favorable attitudes toward products with ethical certifications, however, studies state that there is a large discrepancy between attitudes and actual buying behaviors (De Pelsmacker, Driesen, &

Ray, 2006; Shaw & Clarke, 1999). They suggest that consumers’ attitudes toward products are not reliable predictors of their buying behaviors especially in the domain of ethical consumption. In fact, the demand for socially responsible products remains considerably low; it is reported that these products have less than 1% market share across all categories (MacGillivray, 2000). Considering the awareness of ethical consumerism in the current market, understanding which factors facilitate or inhibit consumers’ purchase of such products may be of theoretical and practical importance. Consumers’ purchase of certain targeted products, in fact, could be increased when marketing strategies are aimed 1 at promoting significant antecedents and removing barriers to the purchase. Additionally, identifying key factors that influence the purchase of socially responsible products would help consumers to be more knowledgeable about their purchase behaviors and enhance their ability to make more informed decision-making.

People seldom make decisions in complete isolation; instead, they are influenced by others in their social environment. Behaviors and expectations of others are taken into account when assessing appropriateness of a certain behavior. In this regard, social norms,

“rules and standards that are understood by members of a group, and that guide and/or constrain social behavior without the force of laws” (Cialdini and Trost 1998, p. 152) are likely to influence consumers’ purchase behavior toward socially responsible products.

Social norms are often employed to promote ethical human behaviors and found to be effective in many cases (e.g., Cialdini, Reno, & Kallgren, 1990; Schultz, Nolan, Cialdini,

Goldstein & Griskevicius, 2007).

Product knowledge also has an important role in explaining consumer behaviors; it generally has a positive influence on purchase decisions (e.g., Brucks, 1985; Park,

Mothersbaugh, & Feick, 1994). Investigating the effect of product knowledge may be especially critical for studies of socially responsible products given the fact that consumer awareness and understanding about these products are reported to be particularly low

(Demeritt, 2002).

1.2. Problem Statement

The present study focused on investigating factors which are presumed to affect socially responsible consumer behavior (SRCB). The major problem which this dissertation intends to examine is whether, and to what extent, a) social norms and b)

2 product knowledge can motivate consumers’ purchase of socially responsible products.

First, the effect of social norms has been widely examined across various types of human behavior. In particular, researchers closely investigated the effect of social norms by distinguishing the concept into two types (i.e., injunctive and descriptive norms) and validated their strong influence on changing one’s ethical behavior such as littering and energy conservation behavior (e.g., Cialdini et al.,, 1990; Schultz et al.,, 2007). Yet, little research has been conducted within the context of ethical consumption. Prior consumer studies in this area have mostly applied a modified version of TRA (theory of reasoned action) or TPB (theory of planned behavior) to examine the influence of subjective or injunctive norms on behavioral intention (Kim, Lee, & Hur, 2012). As a result, the effect of descriptive norms on consumers’ socially responsible purchase behaviors is unclear.

Next, a body of research confirms the strong effect of product knowledge on consumer’s decision-making behavior (e.g., Brucks, 1985; Park et al., 1994). Lack of knowledge is reported to be the main reason that consumers do not buy socially responsible products. For example, more than half (59%) of respondents from a study indicated that they had never considered organic products due to insufficient knowledge about them (Demeritt, 2002). Although product knowledge is considered to be an important determinant of consumer behaviors, the extant literature provides very limited empirical evidence of how it influences SRCB.

1.3. Purpose

The purpose of the present study was to address the aforementioned gaps in the literature and contribute to a better understanding of socially responsible consumers by identifying significant factors influencing their purchase decisions. In order to

3 accomplish this goal and to precisely examine diverse aspects of consumer behaviors in this context, the dissertation consists of three parts: a meta-analysis of the TPB literature in the SRCB context (Study 1), an examination of the effect of social norms and product knowledge on purchase of organic cotton and fair-trade apparel (Study 2), and an examination of the efficacy of using social media to increase knowledge and promote

SRCB (Study 3).

A comprehensive review (Study 1) was conducted of the SRCB literature as a preliminary investigation to examine the effectiveness of the theory of planned behavior

(TPB) that is used in Study 2. Socially responsible consumers, also referred to as pro- social consumers (Osterhus, 1997; Nilsson, 2008) and ethical consumers (De Pelsmacker et al., 2005), have been conceptualized somewhat broadly; the term itself and its definition vary across studies. The complexity of the term’s meaning and mixed results in previous studies provide narrow understanding of consumers’ purchase behaviors, making this segment of consumers difficult to identify. Therefore, the first part of the study (Study 1) reviews previous studies on socially responsible consumers and identifies major antecedents of behavioral intentions to contribute to the theoretical understanding of their purchase behaviors. In particular, the theory of planned behavior (TPB) (Ajzen,

1991) is one of the most dominant theories that have been applied to predict behaviors in the fields of consumer behavior and marketing. Study 1 reviews applications of TPB in the domain of SRCB and verifies its efficacy to explain and predict purchase behaviors.

The influence of social norms on SRCB is a subject of long standing controversy among researchers. Generally social norms are considered to be a uni-dimensional term but several studies (e.g., Cialdini, Kallgren, & Reno, 1991; Reno, Cialdini, & Kallgren,

1993) point out the need to classify the term into two distinctive concepts (i.e., descriptive and injunctive social norms) to appropriately examine the impact on human behavior. These studies found that each type of social norm indicates a distinct resource underlying human motivation and leads to considerably different behavior in an identical setting (e.g., Reno et al., 1993). Thus, Study 2 examined SRCB in the context of the apparel and textile field, particularly focusing on the effect of the two types of social norms by using the TPB framework and including both norms to test their independent effects.

Furthermore, in order to enhance the predictive power of TPB, two types of product knowledge (i.e., objective and subjective knowledge) are included in the model.

Similar to social norms, studies suggest that knowledge should not be treated as a uni- dimensional construct (e.g., Park, et al., 1994; Selnes & Gronhaug, 1986). The inconsistent findings in previous literature regarding the effect of knowledge may be attributed to the way knowledge is measured because the two knowledge types tend to produce significantly different results when explaining consumer behavior. Thus, to test the usefulness of each construct, the independent roles of objective and subjective knowledge in the purchase of socially responsible products are thoroughly investigated.

Study 3 further highlighted the effect of product knowledge on purchase intention and examined the effectiveness of using social media websites to educate and promote

SRCB. A body of research on SRCB emphasizes the importance of behaviors of others in the social environment in effectively changing one’s behavior; many of these studies focus on social norms to promote consumers’ socially responsible purchasing

(Goldstein, Cialdini, & Griskevicius, 2008; Kim et al., 2012). In this sense, social media

5 can be used as a valuable marketing tool to motivate SRCB since it amplifies the power of consumer-to-consumer communication by allowing an individual to communicate with thousands of others quickly and with little effort (Mangold & Faulds, 2009). Numerous retailers are adopting social media websites to increase sales or to enhance their brand image. However, so far, the ways retailers are utilizing these websites in regards to socially responsible products and consumer responses to them remain largely unknown.

A distinctive feature of popular social media websites such as Facebook and Twitter is that different formats of information can be uploaded and presented to the public.

Common representative types could be video, image, text or a combined method (e.g., image and text combined). Another unique attribute of the social media websites is that they present information to viewers and it varies in terms of its quantity and quality.

Identifying which format and what characteristics of information are the most effective in raising awareness about SRCB will help marketers to better serve customers who are interested in SRCB. Therefore, Study 3 tested the effectiveness of using social media websites, particularly focusing on the effect of information format and characteristics, as a way to increase consumers’ knowledge level and to motivate purchase behaviors.

In sum, this dissertation attempts to provide an in-depth understanding of SRCB by focusing on the role of social norms and product knowledge in the purchase of socially responsible products. Significance of this dissertation lies in empirically verifying the strong effect of these factors on SRCB. As noted previously, there is considerable disagreement among researchers regarding the influence of social norms on human behavior (e.g., Sheppard et al., 1988; Van den Putte, 1991). This study aims to address this ambiguity by distinguishing the concept into different types. In terms of

6 product knowledge, evidence of its influence is often cited from the reports of consumer surveys. For instance, a recent global survey showed that nearly all consumers perceived a low level of knowledge regarding environmentally friendly products and a lack of knowledge was reported to be a major barrier to purchasing these products (Manget et al.,

2009). However, these reports are mostly descriptive in nature and the current literature lacks theoretical approaches that address the impact of product knowledge on the purchase process of socially responsible products.

To begin with, Study 1 clarifies the strong association of subjective norms with behavioral intentions in the SRCB context, which directly shows that consumers’ intention to purchase socially responsible products is strongly linked to the normative influence. Study 2 expands the scope of testing the effect of social norms on SRCB by distinguishing two distinct types of social norms, injunctive and descriptive norms. Study

2 also incorporates two types of product knowledge, objective and subjective knowledge, into the model to confirm their independent effects on purchase of these products.

Motivated by the two previous studies that confirm the critical role of social norms and product knowledge, Study 3 further tests the effectiveness of using social media websites as a way to increase the level of product knowledge and promote SRCB.

1.4. Definition of Terms

1. Affective response: shaping of the emotional states (Lavidge & Steiner, 1961). In the

current dissertation, affective response refers to emotions that develop after exposure

to website information regarding socially responsible products.

2. Attitude: ‘‘the degree to which a person has a favorable or unfavorable evaluation or

appraisal of the behavior in question’’ (Ajzen, 1991, p. 188).

3. Behavioral intention: a powerful indicator of the actual behavior. According to TPB,

it is a function of the following three factors: attitude toward performing the behavior,

subjective norms, and perceived behavioral control (Ajzen, 1991). Purchase intention,

a specific case of behavioral intention, refers to the likelihood that an individual will

buy a particular product.

4. Cognitive response: shaping of the mental or intellectual states (Lavidge & Steiner,

1961). In the current dissertation, cognitive response refers to developing knowledge

about socially responsible products.

5. Conative response: “tendency to treat objects as positive or negative goals” (Lavidge

& Steiner, 1961). This is the final stage in the hierarchy of effects model which

involves approach behaviors from the consumer (Smith, Chen, & Yang, 2008). In the

current dissertation, conative response refers to developing intention to purchase

socially responsible products.

6. Descriptive norms: what is commonly done or how others behave which is based on

observations of others’ action in a certain situation (Cialdini, Reno, & Kallgren,

1990).

7. Ethical consumerism: the practice of purchasing products or services that are

produced in a way that reduces social and/or environmental damage (Roberts, 1993).

8. Environmental consciousness: individuals’ awareness about issues impacting the

earth (e.g. pollution, degradation of the environment, and limited resources) (Kim &

Han, 2010).

9. Information quantity: the amount of information that is available to make an accurate

assessment (Keller & Staelin, 1987).

10. Information quality: “the information's inherent usefulness to consumers” in making

an accurate assessment (Keller & Staelin, 1987).

11. Injunctive norms: what ought to be done which reflects the perceived approval by a

certain culture (Kim, et al., 2012).

12. Media richness: “ability of a communication medium to reproduce the information

sent over it” (Lai & Chang, 2011, p566). Rich media are capable of transmitting

information that can “change understanding within a time interval” (Daft & Lengel,

1986, p. 560).

13. Moral norms: one’s awareness of the correctness/ incorrectness associated with

performing a behavior, taking into consideration “personal feelings of …

responsibility to perform, or to refuse to perform, a certain behavior” (Ajzen, 1991,

p.199).

14. Objective knowledge: the accurate and factual information stored in one’s memory.

This construct is measured by some kind of objective test (Flynn & Goldsmith, 1999)

15. Perceived behavioral control (PBC): ‘‘the perceived ease or difficulty of performing

the behavior’’ (Ajzen, 1991, p. 188). It is determined by an individual’s perception of

the presence of resources or opportunities needed to perform the behavior and the

perceived power of the control factor that facilitates the performance of the behavior

(Ajzen, 1991).

16. Self-identity: “the salient part of an actor’s self which relates to a particular

behavior”. It indicates the extent to which an individual sees him/ herself as

satisfying the criteria of any societal role (Conner & Armitage, 1998, p. 1444).

17. Socially responsible consumer: “one who purchases products and services perceived

to have a positive (or less negative) influence on the environment or who patronizes

businesses that attempt to affect related positive social change” (Roberts, 1993, p.

140).

18. Socially responsible product: a product that exhibits “one or several social or

environmental principles which might affect consumer purchase decision” (Bezencon

& Blili, 2010, p. 1306).

19. Subjective knowledge: one’s perceptions of what or how much they know (Selnes &

Gronhaug, 1986)

Acronyms

1. CSR: Corporate Social Responsibility

2. PBC: Perceived behavioral control

3. SRCB: Socially responsible consumer behavior

4. TRA: Theory of reasoned action (Ajzen & Fishbein, 1980)

5. TPB: Theory of planned behavior (Ajzen, 1991)

The second chapter of this dissertation contains Study 1, a meta-analytic review of TPB explaining purchase behaviors of socially responsible products, the third chapter contains Study 2, a study of the effect of social norms and product knowledge on purchase behaviors of organic cotton and fair-trade apparel, and the fourth chapter presents Study 3, a study testing the efficacy of using social media to increase knowledge

10 and promote SRCB. Lastly, the fifth chapter summarizes and integrates the findings from the three studies and provides an overall discussion.

Chapter 2: Study 1- Explaining Socially Responsible Consumer Behavior: A Meta- Analytic Review of Theory of Planned Behavior

2.1. Overview

Despite retailer and consumer interest in ethical consumerism, prior research provides limited knowledge about purchase behaviors of socially responsible consumers.

One notable challenge to summarizing the SRCB literature is that there is a lack of agreement in defining the term, socially responsible consumer among researchers. Some define the term broadly involving diverse types of ethical issues such as environmental protection, support for employment and human rights, and community support (e.g.,

Roberts, 1995; Mohr & Webb, 2005), whereas others define it more narrowly. There are studies that describe socially responsible consumers by simply involving environmental matters (e.g., Antil, 1984). Studies not only vary in how they describe this particular consumer segment but also report inconsistent findings regarding significant antecedents of their purchase behaviors which eventually lead to providing narrow understanding on this topic. For example, previous studies report inconsistent findings regarding environmental consciousness as an antecedent of purchasing environmentally-friendly products. While some studies showed that environmental consciousness leads consumers to make purchases that are more socially responsible (Hustvedt & Dickson, 2009; Lin,

2009), other studies revealed that environmental consciousness and individual’s purchase behavior have a limited relationship (Butler & Francis, 1997; Kim & Damhorst, 1998).

Furthermore, regardless of the fact that an extensive number of studies have applied TPB to explain SRCB, many researchers have questioned the accuracy of the theory in predicting behavior (e.g., O'Keefe, 2002; Ogden, 2003). The explanatory value of subjective norms of TPB has also been a subject of controversy in the literature (e.g.,

Sheppard et al., 1988; Van den Putte, 1991). Therefore, further studies are needed in this area to clarify conflicting findings and enhance our understanding on this topic.

2.2. Problem Statement

Knowledge about socially responsible purchase behaviors is very limited. There are a few studies that systemically review literature on ethical human behaviors, but they were mostly limited to pro-environmental behaviors (e.g., Aertsens, Verbeke, Mondelaars,

& van Huylenbroeck, 2009; Bamberg and Moser, 2007) and not specifically focused on purchase behaviors (e.g., Bamberg and Moser, 2007). Furthermore, gaps exist in the literature regarding product types and behaviors examined; no prior work has aggregated and compared results from consumer studies regarding different types of socially responsible products, such as food versus apparel versus service goods.

In addition to the gaps in the literature, some researchers call for an examination of the theoretical foundations of SRCB studies. The theory of planned behavior (TPB)

(Ajzen, 1991) has been widely applied to predict behaviors in the fields of consumer behavior, including SRCB. Despite its extensive application in previous studies, some researchers question TPB’s assumption that the theory will sufficiently capture theoretical determinants of behaviors or intentions (e.g., O'Keefe, 2002; Ogden, 2003).

Therefore, the current study aims to address these research gaps and to provide a better understanding of SRCB by conducting a meta-analysis of related literature.

Meta-analysis is a statistical procedure used for combining and contrasting results from independent studies. It can identify significant patterns or disagreement among study results and other interesting relationships can be emerged from those study results. In this regard, using meta-analysis procedures can be especially beneficial in clarifying concepts and relationships among meaningful constructs in the field of SRCB.

However, very limited number of meta-analytic research review studies related to purchase behaviors of socially responsible products. A meta-analysis will help provide a fuller picture of SRCB by aggregating results from different studies. The predictive validity of TPB in relation to SRCB can be explained and common findings across diverse study results can be identified from the analysis. Study 1 systemically reviewed applications of TPB in the literature and verified its efficacy in predicting SRCB.

Significant antecedents of purchase intention for products associated with multiple industry domains and ethical issues were examined in order to identify promising additions to the TPB.

2.3. Purpose of Study

Previous research lacks findings that systematically combine and compare diverse study results on SRCB. Additionally, although TPB has been commonly applied to studies that explain ethical human behaviors, further verification regarding its efficacy in predicting consumer behavior is needed, in this particular study domain. Therefore, the purpose of Study 1 is to systematically review previous studies of SRCB that apply TPB or TRA (TPB is an extended version of TRA) to examine the predictive validity of the model and to identify major determinants of consumers’ socially responsible purchase behaviors. As previously explained, these models have been extensively applied in the

14 particular domain of SRCB, making it easier to locate and summarize the study results reported across various research settings. Several research questions are developed below to guide the research process.

Research Question 1. Which TPB predictors (i.e., attitude, subjective norms, and perceived behavioral control) have a stronger relationship with behavioral intention in the context of SRCB?

TPB specifies three determinants of intention to perform a behavior: attitude toward the behavior, subjective norms and perceived behavioral control. This means that an individual will have a strong intention to perform a given behavior when they evaluate the behavior positively, believe that significant others would prefer them to perform the behavior, and perceive that the behavior is not difficult to perform. In this sense, TPB can be regarded as a causal model; the constructs can be represented mathematically as sets of equations or graphically as path diagrams. When employing such causal models, it is important to carefully specify the paths among constructs and determine their magnitude.

The three predictors in the TPB model usually do not have equal power when explaining one’s intention to perform a behavior. Ajzen (1991), the initial developer of the theory, states that “The relative importance of attitude, subjective norms, and PBC in the prediction of intention is expected to vary across behaviors and situations” (p. 188).

However, in general, the effect of subjective norms on intention is found to be less significant than attitude and perceived behavioral control. Several meta-analytic studies of TRA or TPB suggested that subjective norms are the weakest antecedent of behavioral

15 intention (e.g., Sheppard et al., 1988; Van den Putte, 1991) and as a result, some of them deliberately removed the construct from the model (e.g., Sparks, Shepherd, Wieringa, &

Zimmermanns, 1995)

The first research question of Study 1 examines the SRCB literature to see whether the strength of relationships between TPB predictors (i.e., attitude, subjective norms, and PBC) and behavioral intention varies. The relative magnitudes of the relationships between variables are closely examined.

Although studies suggest that the subjective norms component is a rather weak predictor of intention, the present study presumed that it will have an especially strong association with behavioral intention when investigating SRCB compared to other human behaviors. Chang (1998) mentioned that one’s ethical behavior is greatly affected by the social pressure of significant others. Furthermore, a body of research on SRCB highlights the influence of others’ behaviors in effectively changing one’s own behavior; many of these studies emphasize the role of social norms in promoting consumers’ socially responsible behavior (Goldstein et al., 2008; Kim et al., 2012).

Research Question 2. Are there other constructs that significantly contribute to the prediction of behavioral intention in the context of SRCB?

TPB has been applied successfully to various types of human behaviors and previous meta-analyses support the theory’s strong predictive power (e.g., Ajzen, 1991;

Armitage & Conner, 2001). After reviewing several meta-analyses on TPB and TRA,

Rivis and Sheeran (2003) note that attitude and subjective norms (i.e., constructs of TRA)

16 typically account for 33 % to 50 % of the variance in behavioral intention and the increase in the explained variance after adding PBC is 5% to 12%. In spite of such impressive effect sizes (Cohen, 1998), a substantial amount of variance in behavioral intention cannot be accounted for by the TPB components and is left to be explained.

Admitting this limitation of the TPB, many researchers (e.g., Ajzen, 1991; Conner &

Armitage, 1998; Sheppard et al., 1988) suggest including other variables if they capture a large portion of variance in behavioral intention after controlling for the effect of three variables in the TPB model.

For example, moral norms are known to add power to the prediction of both behavioral intention and behavior (Conner & Armitage, 1998). Moral norms could be described as one’s awareness of the moral correctness/incorrectness associated with performing the behavior and take into consideration “personal feelings of … responsibility to perform, or to refuse to perform, a certain behavior” (Ajzen, 1991, p.199). Therefore, moral norms are considered to be distinct from subjective norms because they are feelings related to personal norms, rather than the social pressure that is associated with subjective norms. Studies suggest that moral norms are especially useful when examining behaviors with an ethical or moral dimension (e.g., Conner & Armitage,

1998). Beck and Ajzen (1991) investigated college students’ unethical behaviors such as cheating on an exam and shoplifting and the results showed that moral norms made a significant contribution to the prediction; the concept increased the explained variance in intention by 3% to 6%.

In addition, self-identity has also been frequently incorporated into the TPB model as studies suggest that it is a useful predictor of a particular behavior (Conner &

Armitage, 1998). It can be defined as “the salient part of an actor’s self which relates to a particular behavior” and indicates the extent to which an individual sees him/ herself as satisfying the criteria of any societal role (Conner & Armitage, 1998, p. 1444). Sparks and Shepherd (1992) studied the role of self-identity using the TPB model to predict a consumers’ intention to purchase organic foods and found that self-identity had a direct influence on both attitude and purchase intention. The results suggest that the more likely consumers think of themselves as green consumers, the more positive their attitude and the stronger their purchase intention with regard to organic foods.

Another variable of interest is environmental consciousness. A large number of studies on SRCB are dedicated to finding determinants of consumers’ green purchase behaviors and in such particular cases, environmental consciousness is known to be a major driving force of performing the behavior (Hustvedt & Dickson, 2009). Many people think it is worth buying or paying extra for green products because their production process has a less negative impact on the environment. Although, there seems to be contradictory results (e.g., Butler & Francis, 1997; Kim & Damhorst, 1998), environmental consciousness is used very often in studies that try to explain consumers’ green purchase behaviors including those that incorporate the TPB model. Based on previous research, Study 1 presumed that the three variables (i.e., moral norms, self- identity, and environmental consciousness) will help improve the predictive power of

TPB in explaining purchase intention of socially responsible products.

Research Question 3. Are there any moderators that influence the relationships between

TPB constructs?

The next focus of Study 1 is to identify potential moderators of the relationships between TPB constructs. First, the study tested the effect of product/service type: consumers’ purchase behaviors within the domains of apparel, food, and hotel/tourism industry were compared. Although their study was limited to environmentally friendly products, Manget et al. (2009) showed that SRCB strongly depends on the product's category. Their study suggested that the highest percentage of consumers were willing to pay more for products in the food category; approximately 30% of the consumers were willing to pay a premium of 10 % or more for seafood, fresh meats, and dairy products.

Consumers’ willingness to pay a premium was also fairly high for wearable products which included apparel and shoes. The hotel/tourism industry was not examined in

Manget et al.’s (2009) study, but it is reported that an increasing number of consumers prefer to stay in ethical lodging facilities that follow socially responsible practices (Kim

& Han, 2010). Consumers’ purchase behaviors tend to be category-specific because product/service attributes, outcome benefits, and purchase experiences vary across product/service categories (Carlson, Vincent, Hardesty, & Bearden, 2009; Low & Lamb,

2000). The concept of socially responsible consumption involves a wide range of product/service categories which include tangible goods such as foods and apparel products as well as intangible goods such as services provided in the hotel and tourism industry. However, in terms of SRCB, scholarly attention has been mostly centered on the food industry, especially on organic food, and thus, the number of studies that examine other product domains is limited (Han & Chung, 2014). Because consumer behavior in the context of a certain product cannot be generalized to other cases, it would

19 be important to see how strengths of relationship between TPB constructs differ across product /service categories.

Second, the effect of ethical issues associated with the purchase was examined:

TPB results associated with products that address two types of ethical issues (i.e., environmental protection and support for human rights) were compared. Roberts (1995) criticized research in the marketing domain for treating SRCB as equivalent to eco- friendly consumer behavior and developing consumer profiles and marketing strategies solely based on environmental matters. He further suggested that we need to take into account the two domains of ethical issues (i.e., environmental and general social matters) to represent a full description of SRCB. In Mohr and Webb’s (2005) experimental study, scenarios were created to manipulate the two ethical domains of environmental protection and support for human rights. According to the results, significant differences were found between consumer behaviors regarding these two matters. For example, CSR (corporate social responsibility) activities had a much stronger effect on consumers’ evaluation of a company when the ethical domain was environmental protection than when it was support for human rights. Thus, in Study 1, TPB results associated with products that address two types of ethical issues (i.e., environmental protection and support for human rights) are compared to test the moderating effect of ethical issue.

Lastly, the location of the study is examined as a moderator to see whether culture has an effect on SRCB. It is commonly accepted that people from distinctive cultural backgrounds respond differently to various types of social matters. For example, people from a collectivistic culture generally put more emphasis on group cohesiveness than those from individualistic culture and thus, are highly influenced by subjective norms

(Lee & Green, 1991). In this sense, social influence may play a stronger role during the purchase process of socially responsible products in collectivistic cultures relative to individualistic cultures. There is a need for addressing this type of problem because in recent years, marketers are increasingly focusing on the role of ethnicity and culture in determining consumer behavior (Holland & Gentry, 1999). Based on previous research,

Study 1 compares the TPB results across study locations to examine the effect of culture on SRCB.

2.4. Literature Review

2.4.1. Theory of planned behavior

The Theory of Planned Behavior (TPB) (Ajzen, 1991) is one of the most influential social-psychological models in explaining attitude-behavior relationships

(Armitage and Conner, 2001). Due to its usefulness in identifying key determinants of human behaviors, TPB has been used in diverse research settings to facilitate the understanding of behaviors in which people engage (Bailey, 2006). According to TPB, behavioral intention is a powerful indicator of the actual behavior, meaning the stronger the intention, the more likely should be the performance of the behavior. Behavioral intention is a function of the following three factors: attitude towards the behavior, subjective norms regarding the behavior, and perceived behavioral control related to performing the behavior.

Attitude can be defined as ‘‘the degree to which a person has a favorable or unfavorable evaluation or appraisal of the behavior in question’’ (Ajzen, 1991, p. 188). It is determined by the individual’s belief regarding the attributes of the behavior or the outcomes produced by performing the behavior ( ) and the individual’s evaluation

21 concerning the importance of the outcome of performing the behavior ( ) (Eagry &

Chailen, 1993).

Subjective norms refer to the influence of others in the social environment in regards to performing the behavior. Ajzen (1991) defined the construct as ‘‘the perceived social pressure to perform or not to perform the behavior’’ (p. 188). Subjective norms are determined by two components: normative beliefs ( ) concerning the expectations of important social referents and motivation to comply with them ( ) (Ajzen, 1985).

Social referents commonly include friends, family members, colleagues, and business partners (Hee, 2000).

Perceived behavioral control is defined as ‘‘the perceived ease or difficulty of performing the behavior’’ (Ajzen, 1991, p. 188). In other words, perceived behavioral control represents the perception of how well an individual can control factors that facilitate or constrain the actions needed to execute a specific behavior. It is determined by an individual’s perception of the presence of resources or opportunities needed to perform the behavior (i.e., control beliefs, ) and the perceived power of the control factor ( ) that facilitates the performance of the behavior (Ajzen, 1991).

Predicting behavioral intention within the TPB model can be expressed as the following multiple regression equation:

B~BI = Att· + SN· + PBC· = (Σ ) + (Σ ) + (Σ )

(B= behavior; BI= behavioral intention; Att= attitude; SN= subjective norms, PBC= perceived behavioral control; = behavioral beliefs; = evaluation; = normative

22 beliefs; = motivation to comply; = control beliefs; = perceived power; w1, w2, and w3 represent relative weights)

TPB is an extension of the theory of reasoned action (TRA) (Ajzen, 1985) and the main difference from the TRA model is that the TPB contains an additional predictor (i.e., perceived behavioral control) to explain behavioral intention more effectively. Including this third predictor in the model helps to elucidate that behavioral intentions are not fully volitional. Thus, TPB is regarded as a more comprehensive version of TRA because TRA emphasizes only volitional personal and social factors, while TPB allows us to investigate the influence of a non-volitional component in addition to the prior volitional components on individual’s behavioral intention (Ajzen, 1985; Dean, Raats, & Shepherd,

2008).

TPB has been widely applied by studies in the psychosocial domain (McEachern,

Schroder, Willock, Whitelock, & Mason, 2007). In particular, the theory has been used to enhance the power of predicting a person’s intention to perform socially responsible behavior in various settings. For example, it was applied to the context of unethical software copying (e.g., Chang, 1998), retail employee theft (e.g., Bailey, 2006), and socially responsible consumer decision making regarding the following topics: environmentally friendly consumption (e.g., Han, Hsu, & Sheu, 2010; Kalafatis, Pollard,

East, & Tsogas, 1999; Sparks & Shepherd, 1992), fair trade certification (e.g., Toulouse,

Shiu, & Shaw, 2006), and counterfeits (e.g., Kim & Karpova, 2010; Penz & Stottinger,

2006).

Although TPB is generally considered to be an effective tool in behavioral studies, many researchers have mentioned the limitations of TPB in that it does not provide sufficient accuracy towards the prediction (e.g., O'Keefe, 2002; Ogden, 2003). Ogden

(2003) criticized its conceptual bases by pointing out that some studies that use TPB showed no effect of attitude, subjective norms or perceived behavioral control. However,

Ajzen and Fishbein (2004) clarified that the comparative importance of the three antecedents could vary depending on the type of behavior and population. Ajzen (1991) additionally stated that TPB allows adding other predictors besides those in the TPB framework if they enhance the predictive power of the model. He suggested that TPB is

“open to the inclusion of additional predictors if it can be shown that they capture a significant proportion of the variance in intention or behavior after the theory’s current variables have been taken into account” (P. 199).

Ogden (2003) further criticized the theory in that many of the TPB studies used self-reports instead of objective measures of behaviors. She argued that self-reports of behaviors can be contaminated or biased and, therefore, the results cannot be trusted.

Ajzen and Fishbein (2004) challenged this argument by stating that it is almost impossible to use objective measures of some behaviors, and in many cases using objective measures could be time consuming and expensive. They continued to argue that self-reported measures were found to be quite accurate in certain behavioral domains, such as environmental behaviors (Kaiser, Frick, & Stoll-Kleemann, 2001) or condom use

(Jaccard, McDonald, Wan, Dittus, & Quinlan, 2002),

2.4.2. Socially Responsible Consumers

Webster (1975) defined the socially responsible consumer as one “who takes into account the public consequences of his or her private consumption or who attempts to use his or her purchasing power to bring about social change” (p. 188). He centered the characteristic on the psychological aspect of social involvement, stating that socially conscious consumers should be aware of the problem (e.g., air pollution), should perceive that their individual power can influence the problem, and should play an active role in the community.

Roberts (1993) assumed that two elements operate in the definition of a socially responsible consumer: environmental concern and general social concern. He described this consumer as “one who purchases products and services perceived to have a positive

(or less negative) influence on the environment or who patronizes businesses that attempt to affect related positive social change” (p. 140).

Numerous studies have defined socially responsible consumer behavior by focusing on the notion of corporate social responsibility (CSR) (e.g., Carroll, 1991; Mohr,

Webb, & Harris, 2001). Mohr et al. (2001) examined consumers’ responsive behavior toward CSR and divided respondents into four consumer groups according to the results.

Those in the highest tier were committed to basing much of their purchasing on CSR and were active in learning about social issues and performance of specific firms. The members of this group participated in various socially responsible consumer behaviors that relate to environmental and social issues. Moreover, they believed they possessed the power to influence a firm’s ethical performance. They were willing to change brands and pay more to purchase from companies they considered socially responsible.

Research has suggested that the number of socially responsible consumers is growing. A global survey examining green attitudes and purchase behaviors (Manget et al., 2009) showed that consumer demand for environmentally friendly products is rising; more consumers in 2008 (34%) systemically looked for and often purchased environmentally friendly products than in 2007 (32%). In a more recent study (Cone Inc.,

2013), the majority (93%) of U.S. citizens believed that company’s CSR performance positively influenced the image of the company. This percentage was much higher compared to 84% in 1993 and 85% in 2010.

In spite of the current increase in the number of socially responsible consumers, extant literature offers limited knowledge about them due to the multifaceted nature of the term and varying results in previously reported studies. For instance, while many studies use a broader definition for describing socially responsible consumers, involving various types social issues (e.g., environmental protection, support for employment and human rights, and community support), some studies define the term in a narrower way

(e.g., Antil, 1984: socially responsible consumption in this study only involved environmental matters). Therefore, it is necessary to examine the nature of socially responsible consumers by aggregating and comparing the results across diverse studies and describe these consumers in a more detailed manner.

Furthermore, several studies have suggested that consumers’ attitudes toward socially responsible products are not reliable predictors of their buying behaviors (Bray,

Johns, & Kilburn, 2011; De Pelsmacker et al., 2005) which makes this segment of consumers even more difficult to identify. Many studies have attested to the significant effect of a company's CSR practices on consumer perceptions of the company and its

26 products (e.g., Brown & Dacin, 1997; Ellen, Mohr, & Webb, 2000). They have commonly demonstrated a positive relationship between CSR and consumers' attitude toward the associated company. However, De Pelsmacker et al. (2005) mentioned that there could be a discrepancy between attitudes and actual buying behavior regarding products from socially responsible companies. While some studies have confirmed a significant influence of consumer attitude on behavior (Ferrell & Gresham, 1985; Vitell,

Singhapakdi, & Thomas 2001), others have suggested that attitude alone is a poor predictor of buying behavior (Shaw & Clarke, 1999) in the ethical consumption context.

Thus, despite the notion that ethical consumerism has been evolving over the last few decades, the limited information on this topic derived from previous studies has highlighted the necessity for a deeper understanding on purchase behaviors of socially responsible consumers.

2.5. Methods

Meta-analysis is conducted to address the previously mentioned research questions. Meta-analysis has become a widely used methodology in the social sciences as it is useful in quantitatively aggregating and comparing the results of different studies.

The procedure for performing a meta-analysis is somewhat similar to that of an empirical study (Lipsey & Wilson, 2001): problem formulation, literature review, coding of variables, analysis, and interpretation. The primary difference between an empirical study and meta-analysis is the unit of analysis; in the former, it is generally the subject while in the latter, it is the research study itself.

2.5.1. Sample of Studies

In order to retrieve studies for the review, the present study:

 used the internet search machine Google Scholar and multiple electronic

databases such as ISI Web of Knowledge and EBSCO. Keywords that were used

for the search were TPB (i.e., theory of planned behavior) or TRA (i.e., theory of

reasoned action) plus the terms related to the scope of the study (e.g., socially

responsible consumers, ethical consumers, pro-social consumers, green

consumers).

 manually searched marketing and consumer related journals: Journal of

Consumer Research, Journal of Marketing Research, Clothing and Textile

Research Journal, and Tourism Management.

 examined citations in the located studies.

In the next step, the abstract and method section of each paper were carefully read.

Because TPB provides clear definitions of the theoretical constructs, studies applying this framework needed to use items that properly reflect such definitions to measure the constructs. To further delimit studies for inclusion, the data for Study 1 was limited to those that contain information on consumers’ purchase behavior toward products associated with 2 types of social issues (i.e., environmental protection and support for human rights) within the domain of apparel, food, and hotel/tourism industry. There were several studies that did not identify the product category, investigating consumer’s purchase intention toward, perhaps, a broad spectrum of products. For instance,

Mostafa’s (2007) study investigated the effect of various cultural and psychological factors on one’s green purchase behavior using the TPB model but failed to meet the criteria of product specification that was needed for inclusion. They used items like “I

(1= dislike; 5= like) the idea of purchasing green.” for measuring attitude toward the purchase and “Over the next month, I will consider buying products because they are less polluting.” for measuring intention to purchase. In such cases, the study was excluded for further analysis.

The effect size selected for the present meta-analysis was r, the Pearson product- moment correlation coefficient. Therefore, for the remaining studies, a correlation coefficient between at least two TPB constructs and sample size needed to be reported.

The unit of analysis for meta-analysis can be in different forms which include articles, laboratories, independent hypothesis tests, independent samples, and a combination method (Cooper, 1989). In the present study, the unit of analysis was an independent sample. Accordingly, one study that used an identical sample with another study written by the same authors was deleted for the statistical analysis. As a result, a total of 30 studies reporting results of 33 independent data sets met the selection criteria

(Table 1).

Author(s) Sample Behavior N Country Arvola et al. (2008) Adults aged 18 and over Purchase organic apples and 202 Italy ready-to-cook organic pizza

Arvola et al. (2008) Adults aged 18 and over Purchase organic apples and 270 Finland ready-to-cook organic pizza

Arvola et al. (2008) Adults aged 18 and over Purchase organic apples and 200 U.K. ready-to-cook organic pizza

Anuwichanont, Tourists who have Visit hotels holding 400 Thailand

30 Mechinda, Serirat, experiences staying in hotels environmental concern

Lertwannawit, & Popaijit in Koh Samet, Thailand (2011)

Bissonnette & Contento Seniors from high schools in Purchase organic food 651 U.S.A. (2001) New York City Continued

Table 1. Description of the Sample of Studies and Data Sets

Table 1 continued

Chen & Tung (2014) Taiwanese consumers Stay at a green hotel 559 Taiwan

Dean, Raats, & Adults aged 18 and over Purchase fresh and 281 U.K. Shepherd (2008) processed organic foods

Dowd & Burke (2013) Grocery buyers aged 18 and Purchase sustainably 137 Australia over sourced food

Han & Chung Female consumers over 18 Purchase organic cotton 200 South Korea years apparel

Han & Kim (2010) Consumers who had stayed at Revisit a green hotel 434 U.S.A. a green hotel within the last 6 months

Han, Hsu, & Sheu Adults 18 years or older Stay at a green hotel 428 U.S.A. (2010)

Continued

Table 1 continued

Hauser, Nussbeck, & Grocery buyers aged 18 and Purchase different food 851 Switzerland Jonas (2013) over product (e.g., organic, fair trade, functional foods, fruits and vegetables, etc.)

Honkanen, Verplanken, Adults Purchase organic food 1,283 Norway & Olsen (2006)

Hustvedt (2006) Adults 25 years or older Purchase organic cotton 422 U.S.A. apparel

Kang, Liu, & Kim University students Purchase organic cotton 701 U.S.A., (2013) apparel South Korea, and China

Continued

Table 1 continued

Kim & Han (2010) Adults 20 years or older Pay conventional-hotel 389 U.S.A. prices at a green hotel

Kim, Lee, & Hur (2012) Adults 19 years or older Purchase environmentally 332 U.S.A. friendly apparel

Ko (2012) Young female consumers of Purchase environmentally 234 U.S.A. 18 - 36 years old friendly apparel

Ko (2012) Young female consumers of Purchase environmentally 194 China 33 17 - 38 years old friendly apparel

Lodorfos & Dennis UK residents Purchase organic food 144 U.K. (2008)

Ma, Littrell, & Niehm 18-28 year old female Purchase non-food fair trade 810 U.S.A. (2012) university students products (e.g., apparel and accessories) Continued

Table 1 continued

Michaelidou & Hassan Consumers from the Island of Purchase organic food 222 U.K.

(2008) Arran in Scotland

Robinson & Smith Grocery buyers aged 18 and Purchase sustainably 550 U.S.A.

(2002) over produced foods

Shaw et al. (2006) Subscribers to the UK Avoid sweatshop clothing 794 U.K.

magazine ‘Ethical Consumer’

Shaw & Shiu (2001) Subscribers to the UK Ethical Purchase fair trade grocery 1,472 U.K.

Consumer magazine products

Shaw, Shiu, & Clarke Subscribers to the UK Ethical Purchase fair trade grocery 736 U.K.

(2000) Consumer magazine products

Continued

Table 1 continued

Song, Lee, Kang, & Boo Visitors who attended the Visit environmentally 400 South Korea

(2012) Boryeong Mud Festival in friendly local festival (i.e.,

2010 Boryeong Mud Festival)

Sparks & Shepard UK residents Purchase organically 261 U.K.

(1992) produced vegetables

Tarkiainen & Sundqvist Adults aged 18 and over Purchase organic bread and 200 Finland

35 (2005) flour

Teng, Wu, & Liu (2013) Adults 20 years or older who Stay at a green hotel 258 Taiwan

were willing to stay in a

green hotel

Umberson (2008) Adults 18 years or older Purchase environmentally 213 U.S.A.

friendly apparel

Continued

Table 1 continued

Vermeira & Verbeke Young adults (19–22 years) Purchase sustainable dairy 456 Belgium (2008) products

Zagata (2012) Consumers who purchase Purchase organic food 1054 Czech Republic organic food on regular basis

2.5.2. Data Coding and Data Retrieval

Each study was coded along several dimensions which included (a) additional predictors of behavioral intention beside the ones in TPB (b) type of product/service, (c) ethical issue associated with the purchase, and (d) location of the study. Correlation coefficients involving behavioral intention, attitude, subjective norms, perceived behavioral control, and additional predictors were retrieved from the studies and particular consideration was given to ensure accurate data retrieval. When a study included more than one behavioral intention (e.g., purchase intention of organic pizza and purchase intention of organic tomato) and reported separate correlation coefficient for each behavioral intention, the average of the coefficients was used for the analysis. This process avoids the violation of the independence assumption underlying the validity of meta-analytic studies. Data retrieval was conducted by two independently working coders and the obtained data were checked several times. Disagreement between the coders was resolved after consultation.

2.5.3. Meta-Analytic Strategy

The correlations from each study were first converted using Fisher’s – transformation (Hedges & Olkin’s, 1985). The – transformed correlations were used to calculate an average in which each correlation value is weighted by the inverse of the within-study variance. Then, the average value is back-transformed to give r+ for interpretation purposes.

Further, to examine the homogeneity in the study outcomes, Cochran (1954)’s Q statistics were computed. The Q value follows the distribution with k-1 degrees of freedom (k =the number of studies), under the hypothesis of homogeneity in the effect

37 sizes. Not rejecting the homogeneity hypothesis indicates that the effect size in the population is the same for all studies included in a meta-analysis (Hunter & Schmidt,

2000). Rejecting the homogeneity assumption, on the other hand, shows that there is heterogeneity among study results which implies that each study is from a population that has a different effect size from other studies in the meta-analysis. In this study, weighted mean effect sizes and Q statistics were computed using Comprehensive Meta-Analysis

Software (CMA).

2.6. Results

As seen in Table 2, the data sets tend to be from recent years and the median sample size of the 33 data sets was 400. Slightly less than half (45.5%) of them were from European countries and 30.3% from the U.S.A. In addition, the data sets specified whether consumer’s purchase intention was associated with (1) apparel, hotel/tourism, or food product/service; (2) environmental protection or support for human rights. The results showed that certain research domains apparently received relatively more scholarly attention than other domains. Approximately half (51.5%) of studies were about food products and 84.8% of studies examined purchase behaviors related to environmental protection.

Variable Value

Median year of report 2009.50

Median sample size 400

Location of the study

U.S.A. 10 (30.3%)

Europe 15 (45.5%)

Asia 6 (18.2%)

U.S.A. and Asia combined 1 (3%)

Australia 1 (3%)

Type of product/service

Apparel 9 (27.3%)

Hotel/Tourism 7 (21.2%)

Food 17 (51.5%)

Ethical issue

Environmental protection 28 (84.8%)

Support for human rights 4 (12.1%)

Combined 1 (3%)

Table 2. Descriptive Statistics of the Studies

1. The relative strength of relationship between TPB predictors (i.e., attitude, subjective norms, and PBC) and behavioral intention:

The first research question was assessed by comparing the weighted mean correlation values. As seen in Table 3, mean effect sizes (r+), 95% confidence intervals

(CI), Fail-safe Ns, and Cochran (1954)’s Q statistics were computed for all relationships.

Table 3 also specifies the number of tests (k) that included each specific relationship.

Confidence interval (CI) is useful in showing the degree of precision of the estimated mean effect size. It shows the range of values which is likely to contain an unknown population parameter. For example, a 95% CI of .52-.54 regarding attitude-intention relationship indicates a 95% probability that the mean population effect size will be between those two values. In addition, to determine the robustness of the results, the

Rosenthal’s (1979) fail-safe N is calculated, addressing the concerns for publication bias, also known as the file-drawer effect. The publication bias refers to the tendency to publish studies with significant findings more frequently than those with insignificant findings. Fail-safe N indicates the number of unpublished studies with null results needed to invalidate the observed effect. For example, in this study, the fail-safe N for attitude- intention is 35,971 which means that we would need to include 35,971 studies with non- significant results to disprove the significance of the current meta-analytic computations

(i.e., intention is correlated with attitude). Since it is extremely unlikely that there are

35,971studies with non-significant results that were unable to be located, the mean correlation ( = .53) between attitude-intention may be considered robust.

Relationship k 95 % CI Fail-safe N Q Attitude-Intention 30 .53 .52-.54 35,971 486.58 Subjective Norm-Intention 28 .50 .48-.51 23,550 525.89 PBC-Intention 21 .39 .37-.41 7,367 207.36 Attitude-Subjective Norm 23 .44 .43-.46 13,048 355.61 Attitude-PBC 20 .27 .25-.29 3,195 309.63 Subjective Norm-PBC 20 .30 .29-.32 4,557 271.86 Note. k= number of tests. = correlation weighted by number of subjects. Q values are all significant (p<.001) rejecting the homogeneity assumption.

Table 3. Meta-analysis of the TPB components

Cohen (1998) suggests the following guidelines for interpreting the magnitudes of correlation coefficients: r = .10 (small effect), r = .30 (medium effect) and r = .50 (large effect). As expected, the mean correlations obtained in the present study indicate that the relationships had medium to large effect sizes. The values for all relationships were highly significant and produced narrow 95 % confidence intervals. Particularly, purchase intention correlated most strongly with attitude ( = =.53, p<.0001). Subjective norms

( = .50, p<.0001) also had a strong association with purchase intention and the strength of association was much stronger than the PBC-purchase intention relationship ( = =.39, p<.0001). The difference in the magnitude between the attitude-intention and subjective norms-intention relationships was minimal, which was in contrast to many meta-analytic study results on TPB or TRA (e.g., Sheppard et al., 1988; Van den Putte, 1991). In addition to the strong correlation of subjective norms and purchase intention, subjective norms also correlated moderately strongly with attitude ( = =.44, p<.0001). The fail- safe N of all relationships exceeded the recommended tolerance level of 5k +10 (where k 41 is the number of tests), meaning that the obtained mean correlations shown in Table 3 can be considered to be robust. The homogeneity statistics indicated a large variation in the correlations reported in the data sets (Q values were all significant, p<.001) which encouraged a search for potential moderators. The results of moderator analyses are addressed below in the section for the third research question.

2. Contribution of other constructs to predict intention:

In order to test the extent to which the additional predictors enhance the prediction of purchase intention after TPB predictors have been controlled, a two-step hierarchical regression was performed using the mean correlation matrix (obtained using the methods for effect size calculation) as the input. In the first step, attitude, subjective norms and perceived behavioral control are entered and in the second step, the additional predictor variable is entered into the equation. Beta coefficients and the change in value in behavioral intention are obtained to assess the contribution of the variable above and beyond the effects of the TPB constructs.

Moral Norms. There were 11 independent data sets (k= 11, n= 6935) that included moral norms in the TPB/TRA model. The hierarchical multiple regression revealed that in the first step, all TPB variables had significant beta coefficients accounting for 39.7% of variance in purchase intention (Table 4). When adding moral norms, there was a significant increase in the variance explained in purchase intention ( change= .02,

= 188.63, p<.001). The percentage of variability accounted for by the predictor variables went up from 39.7% to 41.3% after adding moral norms.

Self-identity. Next, a total of 9 studies, each reporting the result of an independent data set (k= 9, n= 5,946), included self-identity in the TPB/TRA model. In the first step, all TPB variables had significant beta coefficients accounting for 32% of variance in purchase intention (Table 5). When adding self-identity, there was a significant increase

in the variance explained in purchase intention ( change= .03, = 266.38, p<.001). The percentage of variability accounted for by the predictor variables went up from 32.0% to 34.9% after adding self-identity.

Environmental consciousness. There were 7 studies, each reporting results of an independent data set (k= 7, n= 3,604) that included environmental consciousness in the

TPB/TRA model. In the first step, all TPB variables had significant beta coefficients accounting for 48.8% of variance in purchase intention (Table 6). When adding environmental consciousness, there was a significant increase in the variance explained in

purchase intention ( change= .02, = 145.56, p<.001). The percentage of variability accounted for by the predictor variables went up from 48.8% to 50.7% after adding environmental consciousness.

Variable Step 1 β Step 2 β Step 1 Attitude .38*** .32*** Subjective Norms .21*** .18*** PBC .28*** .27*** Step 2 Moral Norms - .15*** .397 .413 Model F 1520.46*** 1218.37*** Note. N= 6935. ***p<.001

Table 4. Hierarchical Regression of Intention on TPB and Moral Norms

Variable Step 1 β Step 2 β Step 1 Attitude .40*** .34*** Subjective Norms .19*** .15*** PBC .21*** .19*** Step 2 Self-Identity - .19*** .320 .349 Model F 932.79*** 797.44*** Note. N= 5,946. ***p<.001

Table 5. Hierarchical Regression of Intention on TPB and Self-Identity

Variable Step 1 β Step 2 β Step 1 Attitude .20*** .16*** Subjective Norms .42*** .39*** PBC .26*** .24*** Step 2 Environmental Consciousness - .16*** .488 .507 Model F 1143.86*** 928.74*** Note. N= 3,604. ***p<.001

Table 6. Hierarchical Regression of Intention on TPB and Environmental Consciousness

3. Moderators that influence the relationships between TPB constructs

As seen in the Q statistic results in Table 3, the correlation between TPB constructs varied widely across studies. Hence, there is a need to examine factors that are expected to affect the magnitude of the relationships between TPB constructs. This will lead to reducing the unexplained between-studies variation.

Prior to the meta-analysis, potential moderating factors were identified by observing previous studies using TPB to explain consumer behaviors. Several different correlation matrices were produced by categorizing the data into specific subsets based on these potential moderators. Table 7 shows the values along with the number of data sets and participants for each relationship within the category. Q statistics comparing correlation coefficients between groups were used to test the significance of the difference (Borenstein, Hedges, Higgins, & Rothstein, 2009).

Product/ service type. First, the relationships between TPB constructs are compared across product/service types (i.e., apparel, food, and hotel/tourism). As expected, the strength of relationship between purchase intention and the predictor variables (i.e., attitude, subjective norms and perceived behavioral control) were dependent on product/service types. Q statistics show that the differences in effect size were significant for all three relationships (A-I: Q= 93.07, df=2, p<.001; SN-I: Q= 135.64, df=2, p<.001; PBC-I: Q= 38.59, df=2 , p<.001). For instance, subjective norms were most strongly correlated with purchase intention when the study involved consumers’ purchase of apparel products ( .59). This effect size was much stronger than that of foods

( .40) and hotel/tourism ( .54). In addition, attitude was most strongly correlated with purchase intention in the hotel/tourism domain ( = .61).

Ethical issue type. TPB results between two ethical domains (i.e., environmental protection and support for human rights) are compared. The majority of the data were from studies that examined consumer response toward products associated with environmental protection and very few involved products associated with support for human rights. There were a total of five studies that dealt with the latter product, but most of these studies reported correlation coefficients of only one or two relationships between the TPB constructs.

Considering the small number of data sets used for products associated with support for human rights, the findings need to be interpreted with caution. However, Q statistics revealed that the differences in effect sizes between subgroups were significant in two relationships (A–I: Q= 73.72, df=1, p<.001; SN-I: Q= 41.03, df=1, p<.001) and insignificant in PBC-I relationship (Q= 3.42, df=1, p=.07). Attitude ( = .54) and subjective norms ( = .51) had much stronger associations with purchase intention when the product contributed to environmental protection than when it involved support for human rights (attitude: = .39; subjective norms: = .37).

Study location. The study location is examined as a moderator to see whether culture has an effect on the SRCB. Study 1 presumed that people from different cultural backgrounds will possess distinct viewpoints regarding a wide range of social behaviors, which may also include SRCB.

Although the between-group homogeneity statistic result showed that the difference in effect size of the attitude-intention correlation is significant (Q= 41.78, df=2, p<.001), the mean correlation values were identical for the two locations of U.S.A. and

Asia (U.S.A.: = .58; Asia: = .58). The strength of the attitude-intention correlation

46 was found to be weaker for European countries (Europe: = .48). However, the magnitude of these values were all relatively strong based on Cohen (1998)’s guidelines.

The differences in effect sizes were also significant for subjective norms-intention and PBC-intention relationships (SN-I: Q= 61.04, df=2, p<.001; PBC-I: Q= 40.28, df=2, p<.001). Subjective norms ( = 57) and PBC ( = .52) correlated with purchase intention most strongly in Asia. As expected, in Asia, the effect size of subjective norms was especially strong as evidenced by an additional result. The strength of subjective norms-purchase intention relationship ( = .57) was almost identical to that of attitude- purchase intention relationship ( = .56), which was in contrast to Western countries. In

U.S.A. and Europe, the effect sizes were higher for attitude-purchase intention than subjective norms-purchase intention relationship, consistent with previous meta-analytic studies on TRA or TPB (e.g., Sheppard et al., 1988; Van den Putte, 1991).

A - I SN - I PBC - I Classification Sample size Sample size Sample size k N k N k N Product/Service Type Apparel 7 2774 .58 8 3,106 .59 5 2152 .34 Hotel/Tourism 7 2868 .61 7 2,868 .54 6 2468 .48 Food 16 7498 .48 13 5,142 .40 10 3835 .36

Ethical Issue Environmental protection 28 11,594 .54 26 9,570 .51 19 6,909 .38

48 Support for human rights 3 2,397 .39 2 1,546 .37 2 1,546 .43

Study Location U.S.A. 9 4131 .56 10 4,463 .51 7 3,058 .42 Europe 13 6160 .48 10 3,804 .42 8 3,148 .36 Asia 6 2011 .56 6 1,611 .57 4 1,411 .52 Note. A= Attitude. SN= Subjective norms. PBC= Perceived behavioral control. I= Intention. = correlation weighted by number of subjects. k= number of tests. N= number of subjects.

Table 7. Meta-Analysis of Moderators

A-I SN-I PBC-I Q df p Q df p Q df p Product/ service 93.07 2 <.001 135.64 2 <.001 38.59 2 <.001 type

Ethical issue type 73.72 1 <.001 41.03 1 <.001 3.42 1 .07

Study Location 41.78 2 <.001 61.04 2 <.001 40.28 2 <.001

Note. A= Attitude. SN= Subjective norms. PBC= Perceived behavioral control. I= Intention. Q= between-group homogeneity statistic. df= degree of freedom.

Table 8. Results of Between-Group Homogeneity Statistic

2.7. Discussion

This study used meta-analysis procedures to (a) quantify the relationship between

TPB constructs, (b) examine the contribution of additional predictors to explain behavioral intention, and (c) identify moderators that influence the strength of the relationships between TPB constructs, in the context of SRCB. It is among the first to quantitatively aggregate and compare the results from SRCB studies involving a broad range of product types and ethical issues.

First, the strength of association between behavioral intention and three predictor variables of TPB (i.e., attitude, subjective norms, and perceived behavioral control) is examined by reviewing SRCB studies. As the theory assumes that the relative importance of TPB constructs varies with the context of the behavior and situation (Ajzen, 1991), it would be meaningful to determine which variables have stronger associations with consumers’ intention to purchase socially responsible products. First of all, the results of this meta-analysis provide evidence that TPB provides a good framework for explaining 49

SRCB. Overall, medium to strong sample-weighted mean correlations were found between purchase intention and predictor variables. Previous meta-analytic studies produced similar effect size values in terms of attitude-intention and PBC-intention relationships. For example, Armitage and Conner (2001) conducted a meta-analysis using a database of 185 studies on TPB and the effect size for attitude-intention was = .49

(present study: = .53) and for PBC-intention, it was = .43 (present study: = .39) which are only slightly different from the values from this study. However, they reported a distinct level of correlation for subjective norms-intention: it was = .34 which is much lower than = .50 of this study. Interestingly, the mean correlation of subjective norms-intention in the present study was very similar to that of attitude-intention and greatly surpassed the mean correlation of perceived behavioral control-intention. This is in contrast to many studies that suggest that magnitude of subjective norms is usually much smaller than the other two constructs in explaining behavioral intention. It implies that SRCB may be considerably affected by the social pressure of significant others. In addition to its direct relationship with intention, subjective norms were also indirectly associated with intention via attitude; the magnitude of the attitude-subjective norms relationship was moderately strong ( = .44). Such a finding is in line with previous research that empirically supports the importance of social norms in the socially responsible consumption context (e.g., Kim, Lee, & Hur, 2012; Schultza, Khaziana, &

Zaleskia, 2008). Thus, appealing to social norms could be an effective way to motivate

SRCB. For instance, retailers may consider adopting marketing campaigns that implement normative messages as a means for promoting socially responsible behaviors such as eco-friendly or fair-trade consumption.

Next, this study examined the extent to which additional predictor variables enhance the prediction of intention. The independent effects of moral norms, self-identity, and environmental consciousness on intention were tested after TPB variables had been statistically controlled. Findings show that the three variables captured similar amounts of additional variance: moral norms explained an increment of 2%, self-identity 3%, and environmental concern 2% of additional variance in intention above and beyond the effect of attitude, subjective norms, and PBC. These figures were quite similar to those reported in previous meta-analysis of the TPB (e.g., Conner & Armitage, 1998). The amount of increase was statistically significant in all three cases and the variables had significant effects on purchase intention above and beyond the effect of TPB components.

Overall, it can be suggested that moral norms, self-identity, and environmental consciousness display good predictive validity in explaining socially responsible buying behaviors. The results further imply that SRCB is determined by the interplay of personal and social motives. Moral norms and self-identity were both strong predictors of purchase intention; these variables are conceptually distinct from subjective norms as they are closely related to personal norms which can be described as one’s own values involved in a particular behavior (Conner & Armitage, 1998). In addition, one’s pro- social and self-interest motives were both strong motivators for the purchase; concerns regarding the environment and attitude toward the purchase had strong influences on purchase intention.

Finally, this study found moderating roles for product type, ethical issue, and culture with regard to the TPB component relationships. For instance, the subjective norms-purchase intention relationship was stronger among those who are shopping for

51 apparel products. Because apparel products are socially visible (Robertson, 1970), purchase of these products may be more affected by others’ opinions and expectations than purchase of other products or services which tend to be less socially visible. A body of literature addresses the role of apparel products as nonverbal modes of communication and socially significant symbols. They are often used as tools for judgments of social appropriateness affecting perceptions of the wearer’s societal roles such as competence

(Douglas & Solomon, 1983), task performance (Lapitsky & Smith,1981), and credibility

(O'Neal & Lapitsky, 1991). Hence, considering the strong social-symbolic meaning of apparel products, consumers may rely greatly on group norms during the purchase process as they would be credible indictors of determining which products are socially acceptable or unacceptable.

The subjective norms-purchase intention relationship was also stronger for Asian consumers. Collectivism, which is prominent in many Asian countries places a strong emphasis on cohesion within social groups. Asian consumers having greater awareness of social aspects related to purchases, thus would be more influenced by social pressure when making socially responsible buying decisions, relative to Western consumers.

Approval from their social groups will be an important consideration for them. Given that

TPB results vary across countries, international marketers should pay close attention to cultural factors that may significantly affect their marketing activities.

Furthermore, the PBC-purchase intention relationship was stronger for hotel/tourism services and for Asian consumers. In general, people are more disposed (i.e. intend) to perform behaviors that are considered to be achievable (cf. Bandura, 1997). In other words, resources and opportunities must exist before the behavior can be performed.

The PBC-purchase intention relationship in this study may be linked with product availability. In the case of SRCB, products/ services must be available to consumers before they make the purchase decision. Otherwise, their intention to purchase will be lower, even though they have favorable attitudes towards the purchase and feel strong social pressure to purchase. It is plausible that in situations in which consumers perceive limited accessibility of the products, PBC exerts a stronger impact on their purchase intentions. For instance, Asia is reported to have a relatively smaller market for organic products compared to the U.S. and Europe (FiBl & IFOAM, 2011). This may explain the result of the strong impact of PBC on purchase intention in Asia: Asian consumers’ perceptions about the difficulty of buying socially responsible products strongly influenced them to have lower intentions to purchase. Therefore, in order to motivate the purchase of socially responsible products/services, it is essential that marketers acknowledge the importance of providing consumers with more opportunities for purchasing and removing barriers that may curtail their approach behaviors.

By combining and contrasting results from previous studies, Study 1 aimed to clarify the nature of SRCB. This study encompassed various types of products and ethical issues to address the gap in previous meta-analytic studies which have been centered on pro-environmental behaviors (e.g., Aertsens et al., 2009; Bamberg & Moser, 2007) and offered limited knowledge of consumers' purchase behaviors (e.g., Bamberg & Moser,

2007). The results of this study, based upon 33 independent data sets involving a wide range of purchase domains, provide strong support for the predictive validity of TPB. As a result, they address the concerns among researchers about the theory’s ability to predict human behavior (e.g., O'Keefe, 2002; Ogden, 2003). In addition to examining the

53 strength of relationships between TPB constructs, this study also examined potential moderators of these relationships, namely, product/ service type, ethical issue type and study location. There has been a considerable amount of inconsistency reported in previous SRCB literature regarding the relationship between TPB constructs and the strength of their effects. Identification of these moderators offers a valid explanation about the differences across study results.

While Study 1 revealed moderators of the relationship between TPB constructs, one limitation of using the meta-analytic procedure is that it cannot directly assess the cause behind such moderating effects. Further examination of the moderators of the TPB relationship may be necessary to enhance our understanding of consumer behavior in this context. In particular, this study assumed that the stronger subjective norms-purchase intention relationship among Asian consumers was due to collectivism, the dominant culture of Asia, which encourages group conformity. However, this is simply based on the author’s assumption and more examination is needed to validate such explanation.

Future studies may, for instance, incorporate the individualism-collectivism scale to confirm why the moderating effect occurred.

Another shortcoming of Study 1 is that small sample sizes were used in several parts of the meta-analysis. Most representatively, there were only two or three data sets that reported TPB relationships associated with products supporting human rights. This was in stark contrast to the reported relationships associated with products supporting environmental protection which involved an average of 24 data sets. Although they revealed some interesting findings, the generality of those findings may be unclear. Thus, further comparison between consumers’ purchase behavior toward these two products

54 would be necessary to verify the findings from the current meta-analysis. Because only a limited number of studies have examined the latter ethical issue (i.e., support for human rights), meta-analysis may not be a successful procedure for the comparison. Future studies may assess consumer behavior toward the two products by testing an identical model and comparing how the results differ in order to provide a clearer idea. Inspired by such notion, Study 2 empirically tested a modified TPB model with additional predictors and used multi-group SEM to compare results between products associated with two ethical issues.

Chapter 3: Study 2- The Effect of Social Norms and Product Knowledge on Purchase Behaviors of Organic Cotton and Fair-Trade Apparel

3.1. Overview

Study 1 confirmed that the TPB framework is suitable for and effective in explaining consumers’ intention to perform socially responsible behaviors. Fishbein

(1997) states that, because it describes a causal process, TPB has the power to help design interventions that can produce change in human behavior. Therefore, application of TPB can be especially useful to researchers who seek to identify ways to motivate

SRCB. The problem, however, is that existing SRCB literature has shown that utilizing the model leaves a substantial amount of variance yet to be explained. According to the results of Study 1, TPB predictors accounted for 32% - 49% of variance in purchase intention which is very similar to the 33 % - 50 % range reported in Rivis and Sheeran's study (2003). To further enhance the TPB’s predictive ability, researchers suggest including supplementary variables that capture a large amount of variance in behavior or behavioral intention (Ajzen, 1991; Conner & Armitage, 1998; Sheppard et al., 1988). For example, Rivis and Sheeran (2003) found that descriptive norms increased the explained variance in behavioral intention by 5% after controlling for the effect of attitude, subjective norm, and perceived behavioral control. They suggested that descriptive norms may qualify as a supplementary predictor in the TPB model.

Social norms and product knowledge are often considered as significant factors that influence consumer behaviors. Consumers take into account behaviors and expectations of others when assessing the appropriateness of their purchase behavior.

Employing social norms has been found to be effective in promoting ethical behaviors

(Cialdini et al., 1990; Goldstein et al., 2008) as well as purchase related behaviors (Kim et al., 2012). Also, product knowledge is known to have a positive influence on consumers’ purchase decisions (e.g., Brucks, 1985; Park et al., 1994). Previous literature has shown that greater level of knowledge or awareness concerning socially responsible products positively affects forming attitudes and purchase behaviors (Aertsens,

Mondelaers, Verbeke, Buysse, & Van Huylenbroeck, 2011). Therefore, the current study applies a modified TPB model by adding social norms and product knowledge into the framework.

3.2. Problem Statement

The impact of social norms on behavior is a subject of controversy across studies.

For instance, contrary to some empirical findings that support the inclusion of subjective norms in the theoretical model for predicting behavioral intention, several meta-analyses on TRA or TPB show that they have a very weak influence on behavioral intention (e.g.,

Sheppard et al., 1988; Van den Putte, 1991).There are studies that suggest that this problem may be resolved by classifying social norms into two distinctive types (i.e., descriptive and injunctive social norms), as each type indicates a distinct resource underlying human motivation and leads to different human behaviors (e.g., Cialdini,

Kallgren, & Reno, 1991; Reno, Cialdini, & Kallgren, 1993). SRCB studies that examine the effect of social norms generally apply the TPB framework (the subjective norms

57 construct from the TPB is conceptually identical to injunctive norms) and treat social norms as a uni-dimensional construct, leaving the effect of descriptive norms largely unknown. Since the two types of norms impact behavior differently, there is a need to distinguish the meaning of social norms to appropriately examine their independent impact on SRCB. By thoroughly investigating the independent effect of each norm construct, the results of Study 2 provide valuable managerial insights to marketers who are interested in promoting ethical consumption. The results not only reveal whether norm-focus strategies will be effective in motivating consumers’ purchase, but also, indicate which particular normative elements they should focus on to boost their marketing outcomes.

In addition, investigating the effect of product knowledge may be crucial especially in the ethical consumption domain, as consumers have relatively poor understanding of the related products. From the perspective of practical value, it may be important to discuss the potential of consumer education to motivate purchase of these products. Thus, Study 2 is designed to gain information on how consumers’ knowledge level influences purchase behaviors. Two types of product knowledge are incorporated into the model to examine their effects and to enhance the predictive power of TPB: objective knowledge, a measure of what or how much a person actually knows and subjective knowledge, a self-assessed level of knowledge. Obtaining information on how each type of knowledge operates in the process of making socially responsible purchase decisions will help marketers develop more effective strategies, mainly regarding consumer education. From the consumers’ point of view, understanding the strong influence of product knowledge on their socially responsible purchase decisions may

58 encourage them to seek more learning opportunities related to their purchase from which they may gain skillful judgments to discern reliable and accurate product information from those that are not and hence, become more knowledgeable consumers.

3.3. Purpose of Study

The purpose of Study 2 is to empirically examine major antecedents of consumers’ purchase intention toward socially responsible apparel products. As mentioned previously, a major focus of the study is placed on the effect of social norms and product knowledge on consumers’ decision making.

In particular, Study 2 added social norms and product knowledge to the original

TPB constructs to examine the extent to which they have an effect on SRCB.

Significance of the current study lies in understanding the impact of these variables in predicting consumer attitudes and purchase intentions towards socially responsible products. Marketers often use diverse messages in their persuasive strategies to appeal to consumers. Such messages, including those that reflect normative beliefs or contain product related information, can be either effective or ineffective; some of them can significantly influence recipients’ responsive behaviors. Therefore, identification of the factors that stimulate consumers’ purchasing behavior related to socially responsible products has practical importance to retailer or marketers. Furthermore, the results will help consumers obtain better perception of their buying behaviors. It will give them insight into the role that product knowledge and social influence play when making socially responsible purchase decisions.

3.4. Literature Review

3.4.1. Theory of Planned Behavior

The theory of planned behavior (TPB; Ajzen, 1985) is an influential and a well- supported social psychological theory for explaining human behavior. An important premise of TPB is that an individual’s behavioral decision results from a reasoned process in which the behavior is influenced by attitudes, norms, and perceived behavioral control. Specifically, TPB suggests that attitude (i.e., evaluation of performing behavior), subjective norms (i.e., influence of others on performing the behavior), and perceived behavioral control (i.e., perceived control over performing the behavior) have an impact on the behavior primarily through their influence on the behavioral intention. Therefore, behavioral intention is considered to be a strong determinant of the behavior.

Despite the fact that subjective norms are included in the TPB model as a major component, the role of subjective norms has often been criticized due the relative weakness of the subjective norms-behavioral intention relationship (e.g., Sheppard et al.,

1988; Van den Putte, 1991). It has been suggested that the weak association between the two constructs reflects the idea that intention is mainly affected by personal factors (i.e., attitude and perceived behavioral control) (Ajzen, 1991). On the other hand, there are studies pointing out that the problem may be with the narrow conceptualization of the theory’s normative component, as it may be responsible for lessening the strength of the subjective norm-intention relationship (e.g., Armitage & Conner, 2001).

In cases where the normative component is considered to be an important motivator of decision-making, it is often included in the TPB model as an antecedent of attitude. Although the subjective norms-attitude relationship is not mentioned in the original TPB work, many previous studies have shown that subjective norms have a

60 significant influence on forming positive or negative attitudes (Han et al., 2010; Kalafatis et al., 1999; Kim & Karpova, 2010; Tarkiainen, & Sundqvist, 2005).

3.4.2. Focus Theory of Normative Conduct

The focus theory of normative conduct clarifies the concept of social norms which has a long standing controversy among researchers regarding its explanatory value

(Cialdini et al., 1991; Reno et al., 1993). The theory distinguishes between the concepts of descriptive and injunctive social norms because each indicates a distinct resource of human motivation (Deutsch & Gerard, 1955). Descriptive norms are generally defined as what is commonly done or how others behave which is based on observations of others’ action in a certain situation. It is what the majority do and motivates individuals by giving evidence of what tends to be an adaptive and effective action (Cialdini et al., 1990). On the other hand, injunctive norms could be described as what ought to be done which reflect the perceived approval by a certain culture (Kim, et al., 2012). Rather than simply informing individual's actions, injunctive norms direct behavior based on the promise of social permission.

The application of the focus theory of normative conduct may have substantial practical value by providing suggestions on the type of norms that need to be targeted in interventions that are designed to promote engagement in a certain behavior. There have been numerous programs that have used a normative message as the primary tool for changing social behaviors such as eco-friendly consumption, drug use, disordered eating, littering, and recycling (e.g., Cialdini et al, 1990; Goldstein et al, 2008; Kim et al., 2012).

Kallgren et al. (2000) mentioned that for those who are interested in promoting socially beneficial behaviors through marketing campaigns, using practical implications of this

61 theory could be extremely helpful. For example, companies are increasingly employing socially responsible claims in terms of materials and packaging for promoting their products. It is expected that use of the focus theory of normative conduct will help identify what kinds of cues motivate consumers to respond and purchase socially responsible products.

3.4.3. Product Knowledge

TPB regards knowledge as an antecedent of an individual’s attitude (Eagly &

Chaiken, 1993). Product knowledge is based on one’s understanding or awareness about the product and the level of confidence one feels about it (Lin & Chen, 2006). It reflects an individual’s cognitions which influence different stages of decision-making, including buying decisions. In fact, there has been a large amount of research about the effect of product knowledge on consumer behavior (e.g., Brucks, 1985; Lin & Chen, 2006; Sujan,

1985). Studies suggest that product knowledge is associated with various phases of consumer behavior such as product evaluation, purchase intention, and satisfaction.

Hence, the significant role of product knowledge in consumer behavior creates a strong rationale for incorporating product knowledge into the TPB model for the present study.

It would be especially important to test the effect of knowledge on consumer’s acceptance behavior for socially responsible products because it is reported that, in general, knowledge of such products is considerably low among consumers. Manget et al.’s (2009) study on environmentally friendly products illustrates the importance of product knowledge in this particular domain. Their results showed that nearly all the respondents in the survey indicated that they felt confused when shopping for environmentally friendly products and were unsure about what environmentally friendly

62 means and what benefits it provides. Lack of product knowledge was indeed one of the main barriers to purchase in their study. Consistent with this finding, other studies showed that the knowledge of and familiarity with socially responsible products are low, despite the fact that consumers generally have positive viewpoints toward these products

(Ha-Brookshire & Norum, 2011; Lin, 2009). Yet, when learning opportunities were given to consumers, the level of knowledge as well as attitudes toward the products increased significantly (Ha-Brookshire & Norum, 2011). These studies emphasize the key role of product knowledge on consumer decision making and further highlight the need to educate and explain ethical aspects of socially responsible products to motivate SRCB.

Regarding knowledge, two conceptually different constructs could be distinguished: objective knowledge which is defined as the accurate and objective information stored in one’s memory and subjective knowledge which is defined as one’s perceptions of what or how much they know (Park et al., 1994; Selnes & Gronhaug,

1986). The distinction between the two types is important as each knowledge type is reported to have different effects on consumer behavior (Selnes & Gronhaug, 1986; Feick et al., 1992). While there is a positive and significant relationship between objective and subjective knowledge, the level of association between the two constructs is usually not very strong. Reported correlations between subjective and objective knowledge often fall in the .30 - .60 range (e.g., Feick et al., 1992; Brucks, 1985).

Figure 1. Conceptual Model of Study 2

3.5. Conceptual Model and Hypotheses Development

3.5.1. Effect of Attitude and Perceived Behavioral Control on Purchase Intention

According to the theory of planned behavior (TPB), an individual is more likely to perform a behavior when he/she has positive attitudes toward performing the behavior

(Ajzen, 1991). Previous studies on SRCB have supported the positive and strong association between attitude and purchase intention across diverse industry sectors, including organic food (Sparks & Shepherd, 1992; Dean et al., 2008), green hotels (Kim

& Han, 2010; Han et al., 2010), fair trade products (Ma, Littrell, & Niehm, 2012), and environmentally friendly apparel (Kang & Kim, 2013).

In addition, TPB indicates that a person’s behavioral intention is influenced by perceived behavioral control (PBC) (Ajzen, 1991) which reflects an individual’s perceptions of his/her ability to perform the behavior. The strong association between the

64 two constructs has also been empirically supported in previous SRCB studies (e.g., Dean et al., 2008; Kim & Han, 2010; Han et al., 2010; Ma et al., 2012) (see Figure 1).

H1: Attitude toward purchasing socially responsible products will influence the intention to purchase such products.

H2: Perceived behavioral control over purchasing socially responsible products will influence the intention to purchase such products.

3.5.2. Effect of Social Norms on Attitude and Purchase Intention

TPB (Ajzen, 1991) predicts that social norms will influence individual’s behavioral intention. The significant relationship between social norms and purchase intentions has been supported by many SRCB studies (e.g., Dean et al., 2008; Kim & Han, 2010; Han et al., 2010). The relationship between social norms and attitude, which is not included in the original TPB model is often found to be significant when studying socially responsible consumers (Han et al., 2010; Kalafatis et al., 1999; Kim & Karpova, 2010;

Tarkiainen, & Sundqvist, 2005). That is, social norms had an impact on forming positive or negative attitude toward the purchase. Including these cited studies, the majority of studies that apply the TPB framework usually test for the effect of social influence by including a single social norm construct (i.e., subjective norms, which is identical to injunctive norms in most studies) into the model.

The focus theory of normative conduct (Cialdini et al., 1991), on the other hand, clarifies the concept of social norms by differentiating between the concepts of descriptive norms (i.e., what is commonly done by others) and injunctive norms (i.e., what ought to be done) and indicates that the two norms produce significantly different

65 results. When examining socially responsible behavior, several studies showed that injunctive norms had a stronger influence on the behavior than descriptive norms (Kim et al., 2012; Reno et al., 1993).

H3: Social norms relating to purchase of socially responsible products will influence the attitude toward purchasing such products.

H3.1: Injunctive norms relating to purchase of socially responsible products will influence the attitude toward purchasing such products.

H3.2: Descriptive norms relating to purchase of socially responsible products will influence the attitude toward purchasing such products.

H4: Social norms relating to purchase of socially responsible products will influence the intention to purchase such products.

H4.1: Injunctive norms relating to purchase of socially responsible products will influence the intention to purchase such products.

H4.2: Descriptive norms relating to purchase of socially responsible products will influence the intention to purchase such products.

3.5.3. Effect of Product Knowledge on Attitude and Perceived Behavioral Control

Results from a meta-analysis by Bamberg and Moser (2007) emphasized the role of knowledge and confirmed that knowledge directly influences the degree of attitude and

PBC toward socially responsible behaviors. Consistent with this finding, several SRCB studies investigated the effect of product knowledge on consumers’ purchase decision, and they found that knowledge positively affected both attitude and PBC (Kang, Liu, &

Kim, 2013; McEachern & Warnaby, 2008). In these studies product knowledge usually

66 has a much stronger influence on PBC than attitude. The strong association between knowledge and PBC implies that when consumers have greater product knowledge, they tend to perceive less difficulty in and more control over purchasing the socially responsible products. Generally, previous studies that examine the effect of product knowledge on consumer behavior, including the studies cited above, treat knowledge as a single construct, usually measuring either objective or subjective knowledge.

In order to closely investigate the associations between knowledge and purchase behaviors, Pieniaka, Aertsensa, and Verbeke (2010) distinguished between consumers’ subjective and objective knowledge. They found a significant effect of knowledge on consumers’ purchase behaviors of organic vegetables. To be more precise, subjective knowledge was strongly and directly associated with purchase behaviors while objective knowledge was indirectly related to purchase behaviors through subjective knowledge.

Similarly, Aertsens et al., (2011) found that high level of knowledge led to a greater likelihood for purchasing organic food; both objective and subjective knowledge related to organic food were positively associated with attitudes towards the consumption of organic food. They also found that objective and subjective knowledge were positively correlated: Pearson correlation coefficient was reported to be .50. Although significantly correlated, the results show that two constructs are far from being perfectly correlated which was consistent with previous studies (e.g., Feick et al., 1992; Brucks, 1985). Thus, it is necessary to pay attention to the differences between the two measures, as each would have different influences on consumer behaviors. Noticeably, studies on SRCB are centered on examining purchase behaviors related to foods and there is a lack of

67 empirical research that tests the relationship between different types of knowledge and

SRCB within other product categories.

H5: Product knowledge about socially responsible products will influence the attitude toward purchasing such products.

H5.1: Objective knowledge about socially responsible products will influence the attitude toward purchasing such products.

H5.2: Subjective knowledge about socially responsible products will influence the attitude toward purchasing such products.

H6: Product knowledge about socially responsible products will influence the perceived behavioral control over purchasing such products.

H6.1: Objective knowledge about socially responsible products will influence the perceived behavioral control over purchasing such products.

H6.2: Subjective knowledge about socially responsible products will influence the perceived behavioral control over purchasing such products.

3.5.4. Effect of Ethical Issue Type on Consumer Response

Previous studies show that SRCB depends considerably on the type of ethical issue associated with the purchase. Yet, inconsistent findings are reported in prior studies regarding this matter. Loureiro and Hine (2002) assessed U.S. consumers' preferences for organic, locally-grown, and GMO-free food products and concluded that locally grown resulted in a higher premium than did the other two attributes. In terms of apparel products, Hustvedt and Bernard (2008) found that U.S. university students were willing to pay the highest premium for organic products, which was slightly greater than the

68 premium for GMO-free products. Their preference for locally grown fiber products was much lower than organic or GMO-free products.

In addition, Mohr and Webb’s (2005) study with U.S. adults found significant differences between consumer behaviors regarding two types of ethical issues: environmental protection and support for human rights. That is, CSR (corporate social responsibility) activities had a stronger influence on consumers’ evaluation of the company when they reflected environmental protection rather than support for human rights. Therefore, it is hypothesized in this study that the ethical matter associated with the product will have an impact on consumer responses during the decision making process including perceived knowledge, attitude, PBC, perceived social influence, and purchase intention.

H7: Consumer response will differ based on the type of ethical matter associated with the product (i.e., environmental protection and support for human rights).That is, the means of constructs and structural relationships between constructs will differ between the two ethical matter groups.

3.6. Methods

3.6.1. Identification of Socially Responsible Products

Several reliable scales that measure individual’s level of socially responsible consumption were examined to identify which domains, among various alternatives, representatively characterize ethical consumerism. As seen in Table 9, previous studies used various socially responsible consumption scales that typically cover the following ethical matters: (1) environmental protection, (2) support for employment and human

69 rights, and (3) community support. While some scales measure consumer behavior in a narrow way (e.g., Antil, 1984: socially responsible consumption scale in this study only dealt with environmental matters), other measures involve a wider range of social issues.

Roberts (1995) mentions that the dynamic nature of the term, socially responsible consumption, requires continuous refinement of its scale as our understanding related to the term evolves over time. Nevertheless, there seems to be a common thread among various scales that measure this construct. For example, they commonly include items that inquire about consumers’ response to CSR (corporate social responsibility) concerning environmental protection (Table 9). Most of them had the highest number of items allocated to measure this particular domain. Another major theme that the scales generally include is consumer response to company’s philanthropic activities, most noticeably improving employment/ human rights conditions. There were a few that included CSR activities that involve supporting local communities or charities, however, the number of such items seemed to be much less compared to that from the other two domains. Thus, two major themes were selected for Study 2 (i.e., 1. environmental protection and 2. support for employment and human rights) to further test the hypotheses for Study 2.

Study Scale name Number Scale measures consumer of items response to CSR regarding the domain(s) of… Antil (1984) Socially Responsible 40 Environmental protection Consumption Behavior Scale

Frederick & Socially Conscious 8 Environmental protection, Webster (1975) Consumer Index Employment and human rights (SCCI)

Roberts (1995) Socially Responsible 25 Environmental protection, Consumer Behavior Employment and human rights, (SRCB) Scale Community support, Socially responsible marketing

Mohr & Webb Socially Responsible 26 Environmental protection, (2005) Purchase and Employment and human rights, Disposal Scale Community support, Animal rights

Table 9. Domains of Scales that Measure Individual’s Level of Socially Responsible Consumption

The present study focused on SRCB regarding apparel products. In general, scholarly attention in the SRCB literature has focused mostly on food products, with little research examining other product categories. As the results of the meta-analysis show

(Study 1), SRCB strongly depends on the product category and thus, results cannot be generalized to different categories. Although a growing number of apparel studies have recently examined consumer behavior on this particular topic, these studies provide insufficient explanation of how variations in types of social norms and product knowledge impact behavior. Furthermore, as seen in the meta-analysis (Study 1) results, apparel studies that examine SRCB are typically focused on environmentally friendly

71 behaviors (e.g., organic apparel purchase) and lack information on consumer behavior regarding products within other ethical domains including those that reflect support for employment and human rights (e.g., fair trade apparel purchase). Direct comparison between consumers’ purchase behaviors associated with two major ethical issues (i.e., 1. environmental protection and 2. support for employment and human rights) were made in

Study 2, providing a broader understanding of SRCB. The annual retail sales of clothing and accessories in the U.S. was reported to be approximately $223 billion in 2011 (US

Census Bureau, 2014), a figure that directly illustrates the significance of the apparel industry sector.

A preliminary survey was designed to identify the apparel products consumers typically associate with each of the two domains. There were 6 questions in total; for each domain, respondents were given two multiple choice questions and one open-ended question (Appendix A). The three items are as follows: a) indicate the extent to which they believe the ethical issue would be an important concern to socially responsible consumers, b) choose from a list of apparel products they most relate to as supporting the ethical issue, and c) directly type the name of the product if the answer they had in mind was not shown in the list of apparel products from the previous multiple choice question.

After getting the IRB approval, a preliminary survey (Appendix A) was conducted via Amazon Mechanical Turk (MTurk). Amazon MTurk is an online platform for recruiting subjects to perform tasks. To initiate a survey in MTurk, a researcher establishes an account and posts a listing that describes the study and the rewards for taking the survey. An individual (MTurk “Worker”) can review the list of studies and choose to undertake any surveys for which he/she is eligible.

A total of 50 respondents were recruited to identify apparel products consumers typically associate with the two ethical domains of 1) environmental protection and 2) support for employment and human rights. This preliminary study was exploratory in nature, identifying variables required for the subsequent quantitative study. When applying a qualitative approach, Charmaz (2006) claimed that a sample size of 25 is often sufficient for smaller projects and Ritchie, Lewis, and Elam (2003) mentioned that qualitative study samples generally lie under 50.

Data were collected from 50 adult consumers residing in the U.S. First, regarding the ethical domain of environmental protection, 96% of the respondents indicated that purchasing and using eco-friendly products that minimize damage to the environment is an important concern for socially responsible consumers. The questionnaire inquired about the product they most relate to as supporting environmental protection in the fashion and apparel industry; 46% answered organic cotton, 42% recycled cotton, 8% recycled polyester clothing, and 4% bamboo fabric clothing. Thus, organic cotton apparel, which was selected by the highest proportion of respondents as representative of the environmental domain, was selected for the main study.

Next, regarding the ethical domain of support for employment and human rights,

88% of the respondents agreed that purchasing and using products that support fair labor practices and human rights would be an important concern for socially responsible consumers. When asked about the product they most related to for supporting the improvement of employment and human rights conditions in the fashion and apparel industry, 38% of them answered fair trade apparel, 30% child labor free apparel, 24% sweatshop free apparel, and 8% union made apparel. Fair-trade apparel, which was

73 selected by the highest proportion of respondents as representative of the employment and human rights domain, was selected for the main study.

3.6.2. Instruments

A web-based questionnaire in which respondents self-report their answers was developed to measure the constructs used for the testing of the hypotheses. All questions were rated on a 7 point Likert scale (1=strongly disagree, 7=strongly agree) excluding true/false response questions that measured objective knowledge (Appendix B).

 Objective and Subjective Knowledge- Consistent with previous research, the objective knowledge test was developed from a review of several objective sources that provide detailed information on the subject; question answers consist of true/false responses (Fair

Trade International, 2013; Fair Trade USA, 2013; Textile Exchange, 2013). Similar to

Aertsens, Mondelaers et al.’s (2011) measure of objective knowledge, respondents indicated how certain they are of each of their true/ false response on a 5 point scale (1= uncertain, 5= certain) to take into account participant guessing. Respondents who are guessing or, in other words, less certain of the answer received lower scores than those who are more certain of the answer. Following Aertsens et al (2011)’s suggestion, the objective knowledge score was calculated as follows: “ a wrong answer with a certainty of 5, resulted in a score of 0; a wrong answer with a certainty of 4, in a score of 1; a wrong answer with a certainty of 3 in a score of 2, and so on; a correct answer with a certainty of 1 resulted in a score of 5; a correct answer with a certainty of 2, in a score of

6, and so on. The maximum score is thus given to a correct answer with a certainty of 5, which results in a score of 9. The total objective knowledge score is then calculated by

74 summing the scores on each of the four statements and these therefore range between 0 and 36” (p.1360). Aertsens et al (2011) calculated the Cronbach’s alpha coefficient which was .61, indicating a sufficient level of consistency among the four items.

To measure participant’s subjective knowledge, a scale developed by Flynn and

Goldsmith (1999) was used. It included statements like “When it comes to _____, I really don’t know a lot”. Reliability of the scale was assessed by Cronbach’s alpha across three product domains (i.s., clothing, movies, and music) which ranged from .88 to .94.

 Attitude, Perceived Behavioral Control (PBC), and Purchase intention- Attitudes toward purchasing socially responsible products was measured directly by asking respondents to answer several seven-point evaluative semantic differential items. The endpoints of these scales were, for example, “Negative/Positive” and

“Undesirable/Desirable” (Bansal & Taylor, 2002; Conner, Warren, Close, & Sparks,

1999). Cronbach's alphas for the attitude scale were .93 in Bansal and Taylor’s (2002) study and .75 in Conner et al.’s (1999) study. The items for PBC captured respondents’ perceived ability to purchase socially responsible apparel products which included items like “I believe that I have the resources and the ability to purchase _____ apparel”

(Bansal & Taylor, 2002; Conner et al.1999). Cronbach's alphas for the PBC scale were .87 in Bansal and Taylor’s (2002) study and .72 in Conner et al.’s (1999) study. To assess purchase intentions, the respondents indicated on a seven-point scale the extent to which they intend to purchase the socially responsible product. It included items like “I would like to purchase _____ apparel in the future.” (Conner et al. 1999; Kang, Liu, &

Kim, 2013). Cronbach's alpha for the intention scale was .75 in Conner et al.’s (1999) study and .87 in Kang et al.’s (2013) study.

 Injunctive Norms and Descriptive Norms- Injunctive norms were measured with items adapted from Bansal and Taylor (2002) and Fitzmaurice’s (2005) study which used established measures from prior research. They used items like “People who influence my decisions would approve of me buying ____apparel”. Cronbach's alpha scores for the injunctive norms scale were reported to be .85 and above .87 for Bansal and Taylor (2002) and Fitzmaurice’s (2005) study respectively. Items that measure descriptive norms were from Smith et al.’s (2008) study which included questions like “How many of the people who are important to you would buy ____apparel in the near future?”. In this study,

Cronbach's alpha for the descriptive norms scale was .77.

3.6.3. Procedure

Participants were randomly assigned to one of two groups regarding products that reflect 1) environmental protection (organic cotton) and 2) support for employment and human rights (fair trade). After completing the main sections of the survey, they were asked to provide their demographic information such as age, gender, occupation, household income, and previous experience with the product.

3.6.4. Sample

For the main survey, a total of 500 participants were recruited to examine consumer behaviors toward the two products that were identified in the preliminary survey. They were randomly assigned to one of the two product types. Boomsma (1982) used the ratio r= p/ k (p= number of indicators, k= number of latent variables) as a basis

76 to calculate sample size in SEM estimation and suggested having a sample size of at least

100 when r= 4; and at least 400 when r= 2 for adequate analysis. Similarly, Marsh, Hau,

Balla, and Grayson (1988) ran simulations and suggested having a sample size of at least

200 when r= 3; and at least 400 when r= 2. The following equation is based on consolidation and summarization of their results and provides suggestions for determining the lower bound of sample sizes for SEM estimations: n 50 - 450r +

1100 (n= minimum sample size, r= p/ k) . This study used 250 participants for each condition which satisfies the recommended minimum number in the aforementioned studies (r= 3.7 in the present study).

Similar to the preliminary study, participants were recruited from MTurk, a recruitment platform that has been validated by scholars (e.g., Berinsky, Huber, & Lenz,

2012). Ferber (1977) once criticized the over-reliance on university student samples in consumer studies because their results cannot be generalized to a broader population of consumers. Regarding demographics, MTurk participants appear to be more representative of actual consumers than university undergraduates (Berinsky, Huber, &

Lenz, 2012). The present study only used a Generation Y sample (people who are born from 1977 to 1993, Schewe & Noble, 2000) because this generational cohort will make up a major segment of socially responsible consumers that retailers might target.

Generation Y people are recognized as the most consumption oriented among all generations (Sullivan & Heitmeyer 2008), and tend to respond positively to social and environmental issues (Sheahan, 2005).

Table 10 shows the demographic characteristics of the main study sample. More than half of the respondents were male (64.4%), white (76.4%) and had a bachelor’s

77 degree or higher (50.4%). The mean age was 27.96 with a standard deviation of 4.64. In order to ensure homogeneity of the two socially responsible product groups, demographic characteristics were compared to see whether they differ significantly. The results showed that the two groups were not significantly different regarding the following demographic variables: age, t(498) = .86, p = .39; income, t(498) = 1.47, p = .14; education level, t(498) = .35, p = .73; gender, (1, N = 500) = .56, p = .46; race, (5, N

= 500) = .81, p = .98.

While it is not directly comparable with the data from this study which only used the Generation Y sample, a recent U.S. census data (US Census Bureau, 2013) showed that 50.8% of the population were female, 77.7% were white, 28.5% had a bachelor’s degree or higher, and the median personal income was $28,051. The current study’s sample and the U.S. population shared similar characteristics with the exception of gender composition and education level. The sample appeared to have more male respondents and be more educated compared to the U.S. population.

A comprehensive survey on Generation Y (Pew Research Center, 2010) indicated that 61% of the people from this generation cohort were non-Hispanic whites and 54% of them had some college or higher educational attainment. The survey suggested that Generation Y cohort is more ethnically diverse than any other generation cohorts. The percentage of the population for the black and Hispanic were 11% and 10% respectively among Baby Boomers (born from1946 to1964), whereas the percentages were 13% and 19% respectively among Generation Y people. Compared with this information of Generation Y, this study’s sample consisted of more white participants and less black and Hispanic participants with overall higher educational attainment.

Variable Frequency % Gender Male 322 64.4 Female 178 35.6 Age 20-24 136 27.2 25-29 192 38.4 30-34 118 23.6 35-37 54 10.8 Race White 382 76.4 Black or African American 26 5.2 Hispanic or Latino 32 6.4 American Indian or Alaska Native 3 .6 Asian 52 10.4 Other 5 1.0 Annual personal income Less than $10,000 101 20.2 $10,000 - $24,999 117 23.4 $25,000 - $34,999 98 19.6 $35,000 - $49,999 93 18.6 $50,000 to $74,999 64 12.8 More than $75,000 27 5.4 Education Did Not Complete High School 3 .6 High School/GED 50 10.0 Some College 154 30.8 Associate Degree 40 8.0 Bachelor's Degree 207 41.4 Master's Degree 33 6.6 Advanced Graduate/Professional work 12 2.4 or Ph.D.

Table 10. Demographic Summary Statistics of the Sample (n=500)

3.6.5. Data Analysis

The data were analyzed using SPSS 19 and AMOS 21. Hypotheses were tested by structural equation modeling (SEM), using maximum likelihood estimation. The cause- effect relationships among the latent constructs were tested according to the hypothesized model. Further, in order to compare the results between the two product groups, ANOVA and multi-group SEM analysis were used.

3.7. Results

3.7.1. Measurement Model

Confirmatory factor analysis (CFA) addresses the overall model fit and the degree to which the latent variables are reflected by the observed variables (Anderson & Gerbing,

1988). Results of the CFA for each of the socially responsible product groups showed a reasonable fit of the model to the data ( /df = 2.10, p<.01, CFI= .95, TLI= .95,

RMSEA= .07 for the organic cotton apparel group; /df = 2.10, p<.01, CFI= .95,

TLI= .94, RMSEA= .07. for the fair-trade apparel group). Hair, Black, Babin, and

Anderson (2010) suggested that values of the goodness of fit indices required for reasonable fit are /df <3, CFI>.90, TLI>.90 and RMSEA<.08.

Convergent validity and discriminant validity which are considered subcategories of construct validity were also assessed. Table 11 shows item reliabilities, composite reliabilities, and AVEs for the constructs of the measurement model. All measurements achieved a satisfactory level of reliability; item and composite reliability estimates both ranged from .88 to .95. and from .88 to. 94 for organic cotton apparel group and fair-trade apparel group respectively. For both groups, AVE of the constructs exceeded Fornell and

Larcker’s (1981) recommended threshold of .50. Standardized confirmatory factor

80 loadings were all significant (p < .001) and exceeded .60, displaying strong convergent validity (Anderson & Gerbing, 1988). Additionally, as shown in Table 12, AVEs were greater than the squared correlations of the constructs, showing discriminant validity

(Fornell & Larcker, 1981).

Construct Items Organic cotton apparel Fair-trade apparel Factor α Composite AVE Factor α Composite AV Loadings Reliability Loadings Reliability E Attitude .95 .95 .78 .94 .94 .77 For me buying organic cotton (fair trade) apparel would be:

Negative/ Positive .89 .86

Unpleasant/ Pleasant .86 .89

Foolish/ Wise .88 .88

A bad idea/ A good idea .92 .90

Undesirable/ Desirable .88 .87

PBC .93 .93 .76 .90 .91 .71 I believe that I have the .90 .88 resources and the ability to purchase organic cotton (fair trade) apparel.

Table 11. Results of the Final Measurement Models for the Two Groups Continued

Table 11 continued

I do not face high barriers .89 .86 in purchasing organic cotton (fair trade) apparel.

If I wanted to, I could .95 .92 easily buy organic cotton (fair trade) apparel.

How much control do you .75 .69 think you have over purchasing organic cotton (fair trade) apparel in the 83 near future?

Injunctive .88 .88 .72 .90 .90 .76 norms People who influence my .85 .86 decisions would approve of me buying organic cotton (fair trade) apparel. Continued

Table 11 continued

People who are important .88 .89 in my life would approve of me buying organic cotton (fair trade) apparel.

Close friends and family .81 .86 think it is a good idea for me to purchase organic cotton (fair trade) apparel.

Descriptive .88 .88 .72 .88 .88 .72 norms How many of the people .88 .89 84 who are important to you

would buy organic cotton (fair trade) apparel in the near future?

What proportion of the .88 .88 people who are important to you buy organic cotton (fair trade) apparel?

How likely is it that .79 .77 people who are important to you buy organic cotton (fair trade) apparel? Continued

Table 11 continued

Purchase .91 .92 .79 .91 .91 .77 Intentions I would like to purchase .84 .84 organic cotton (fair trade) apparel in the future.

If I see organic cotton .88 .90 (fair trade) apparel, I intend to purchase or consider purchasing it.

If I see a retail store .94 .89 85 selling organic cotton (fair trade) apparel, I intend to visit the store to purchase a product.

Subjective .95 .95 .81 .94 .94 .81 Knowledge I know a lot about organic .95 .93 cotton (fair trade) apparel.

I feel knowledgeable .96 .96 about organic cotton (fair trade) apparel. Continued

Table 11 continued

Among my circle of .83 .79 friends, I’m one of the “experts” on organic cotton (fair trade) apparel.

Compared to most other .86 .91 people, I know more about organic cotton (fair trade) apparel.

Organic cotton apparel (N=250) Fair-trade apparel (N=250) 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1. Attitude .78 .07 .44 .18 .44 .06 .10 .77 .01 .46 .11 .40 .05 .08 2. PBC .27 .76 .04 .08 .13 .00 .01 .08 .71 .04 .08 .06 .01 .01 3. Injunctive norms .66 .21 .72 .24 .31 .05 .07 .68 .21 .76 .30 .40 .03 .10 4. Descriptive norms .42 .28 .49 .72 .36 .02 .26 .33 .28 .55 .72 .23 .02 .07 5. Purchase Intentions .66 .36 .56 .60 .79 .05 .17 .63 .24 .63 .48 .77 .02 .14 6. Objective knowledge .24 .01 .22 .15 .22 - .00 .22 -.11 .18 -.13 .15 - .00 7.Subjective knowledge .32 .08 .25 .51 .41 .05 .81 .28 .12 .32 .26 .37 -.03 .81 Note. Diagonal elements are AVEs, values below the diagonal elements are correlation coefficients, and values above the diagonal elements are squared correlations. AVE of objective knowledge not recorded because it is a composite score.

Table 12. AVEs, Correlations, and Squared Correlations of the Measurement Models

Note. factor loadings from organic cotton apparel data (fair-trade apparel data)

Figure 2. Confirmatory Factor Analysis Model of Study 2

3.7.2. Structural Model

The initial SEM results of the proposed model indicated that the model had a mediocre fit to the data (combined /df= 3.87, p<.01, CFI= .93, TLI= .92, RMSEA = .08; organic cotton /df= 2.69, p<.01, CFI= .93, TLI= .92, RMSEA = .09; fair-trade /df=

2.60, p<.01, CFI= .93, TLI= .92, RMSEA = .08). Thus, an additional relationship was included in the model as suggested by the modification indices (MIs): a correlation between injunctive and descriptive norms. It seemed logical that prevalence of SRCB of significant others (i.e., descriptive norms) and perceived social approval or disapproval of

SRCB (i.e., injunctive norms) would be closely linked. In fact, despite the fact that a large body of research (e.g., Cialdini et al., 1991; Kim et al., 2012; Reno et al., 1993;

Rivis & Sheeran, 2003) suggest that the two types of social norms are distinct constructs which are not interchangeable, many studies show a significant association between them.

For example, in a meta-analysis (Rivis & Sheeran, 2003) examining 18 articles that included TPB constructs and descriptive norms, a moderate level of correlation

(r= .38) was found between injunctive and descriptive norms. In the present study, the correlation was higher than this meta-analysis; r= .52 using the combined data. Therefore, based on these considerations, the initial model was revised letting injunctive and descriptive norms correlate with each other. The fit of the revised model to the data was improved compared to the initial model and was within the threshold of acceptable fit

(combined /df= 2.83, CFI = .96, TLI= .95, RMSEA = .06; organic cotton /df= 2.12,

CFI = .95, TLI= .94, RMSEA = .05; fair-trade /df= 2.07, CFI = .95, TLI= .94, RMSEA

= .07).

3.7.3. Hypotheses Tests

As seen in Table 13, the two product groups yielded similar results in hypotheses testing. For both groups, injunctive norms (organic cotton apparel, β= .60, p<.001; fair- trade apparel, β= .66 p<.001), objective knowledge (organic cotton apparel, β= .12, p= .02; fair-trade apparel, β= .14 p= .01), subjective knowledge (organic cotton apparel,

β= .18, p<.001; fair-trade apparel, β= .13 p= .02) had significant effects on attitude, while descriptive norms (organic cotton apparel, β= .05, p= .47; fair-trade apparel, β= -.04 p= .52) did not. Thus, the results showed consistency between the two groups of supporting H3.1, H5.1, and H5.2 and not supporting H3.2. The SMC (squared multiple correlation) of attitude was .43, meaning that 43% of the variance in attitude was explained by the predictors.

Although objective and subjective knowledge had strong influences on attitude, they did not exert significant effects on PBC as this study initially hypothesized; organic cotton apparel, β= .01, p= .86 (objective knowledge), β= .08, p= .22 (subjective knowledge); fair-trade apparel, β= -.10, p= .12 (objective knowledge), β= .12, p= .08

(subjective knowledge). Consistency in the results between the groups was also found regarding the knowledge-PBC relationship; H6.1 and H6.2 were not supported from the results of both groups.

In addition, for both groups, attitude (organic cotton apparel, β= .45, p<.001; fair- trade apparel, β= .42, p<.001), PBC (organic cotton apparel, β= .15, p= .002; fair-trade apparel, β= .11, p= .03), and descriptive norms (organic cotton apparel, β= .36, p<.001; fair-trade apparel, β= .20 p= .003) had significant effects on purchase intentions,

90 commonly supporting H1, H2, and H4.2. The SMC of purchase intentions was .52, indicating that 52% of the variance in purchase intentions was explained by the predictors.

Interestingly, injunctive norms had a strong and significant influence on purchase intentions for the fair-trade apparel group (β= .21, p= .01), however, such influence was insignificant for the organic cotton apparel group (β= .07, p= .35). Therefore, inconsistency between the two groups was found regarding the injunctive norms- purchase intentions relationship; While H4.1 was supported for the fair-trade apparel group, it was not supported for the organic cotton apparel group.

In conclusion, the statistical results showed that H1 (attitude  purchase intentions), H2 (PBC  purchase intentions), H3.1 (injunctive norms  attitude), H4.2

(descriptive norms purchase intentions), H5.1 (objective knowledge  attitude), H5.2

(subjective knowledge  attitude) were strongly supported, whereas H3.2 (descriptive norms  attitude), H6.1 (objective knowledge  PBC), H6.2 (subjective knowledge 

PBC) were not supported in the data from both groups. H4.1 (injunctive norms  purchase intention) was only partially supported because the proposed path was significant in fair-trade apparel group data and but not in organic cotton apparel group data (Figure 3).

To further test the hypothesis that consumer response will differ based on the type of ethical matter associated with the product (H7), the arithmetic mean of the raw scale scores of attitude, PBC, injunctive norms, descriptive norms, purchase intention, and subjective knowledge were first compared between the two groups. Independent t-tests were performed to see whether the scores differ significantly. As seen in Table 14, scores for descriptive norms, t(498)= -2.71, p= .01; purchase intentions, t(498)= -2.34, p= .02;

91 and subjective knowledge, t(498)= -2.72, p= .01, were significantly higher for the fair- trade apparel group than for the organic cotton apparel group. That is, consumers in the fair trade group 1) perceive that significant others are more likely to buy, 2) have higher intentions to purchase, and 3) believe they know more about the product relative to consumers in the organic cotton apparel group. These results provide preliminary support for H7.

Additionally, multi-group SEM was conducted to determine whether the structural relationships between constructs significantly differed across the organic cotton and fair-trade apparel groups. In other words, the moderating effect of ethical issue was tested. Table 15 reports the results from the datasets of the two distinct socially responsible product groups. To begin with, an unconstrained structural model was examined, in which all path estimates were set to vary between groups. Next, a fully constrained model was estimated in which all structural paths were set to be invariant.

The difference in the chi-square statistics between the two models was insignificant:

= 23.31, p=.62. Although the overall difference test indicated that the causal links in the structural model did not differ significantly, individual structural paths were compared between the two groups to closely examine the moderating effect of ethical issue. The baseline model (i.e., unconstrained model) and a one-at-a-time constrained model in which one structural path at a time was fixed to be invariant (Table 15) were compared by performing the chi-square difference test. Significant difference in the chi- square was found for only one structural path: descriptive norms purchase intentions (p

= .04). As Table 13 shows, this path was significantly stronger for the organic cotton apparel group (β=.36, p<.001) compared to the fair-trade apparel group (β=.20, p=.003).

This means that people are more likely to be influenced by significant others’ purchase behaviors when making their own purchase decisions for organic cotton apparel. Other moderating effects of ethical issue that affected the strength of a relationship between the constructs were not observed.

All Organic cotton apparel Fair-trade apparel

β Result β Result β Result H1: Attitude Purchase Intentions .45*** supported .45*** supported .42*** supported

H2: PBC Purchase Intentions .13*** supported .15** supported .11* supported

H3.1: Injunctive norms Attitude .62*** supported .60*** supported .66*** supported

H3.2: Descriptive norms Attitude .01 not supported .05 not supported -.04 not supported

H4.1: Injunctive norms Purchase Intentions .13* supported .07 not supported .21* supported

H4.2: Descriptive norms Purchase Intentions .30*** supported .36*** supported .20** supported

H5.1: Objective knowledge Attitude .13*** supported .12* supported .14* supported

H5.2: Subjective knowledgeAttitude .16*** supported .18*** supported .13* supported

H6.1: Objective knowledge PBC -.001 not supported .01 not supported -.10 not supported

H6.2: Subjective knowledge PBC .09 not supported .08 not supported .12 not supported Note. ***p<.001, **p<.01, *p<.05

Table 13. Comparison of Path Coefficients

Organic cotton Faire-trade Constructs difference t-value df apparela(SD) apparela(SD)

Attitude 5.33(1.15) 5.42(1.05) -.09 -.86 498

PBC 4.74(1.45) 4.51(1.33) .23 1.81 498

Injunctive norms 4.63(1.05) 4.80(1.08) -.18 -1.87 498

Descriptive norms 2.99(1.26) 3.29(1.23) -.30** -2.71** 498

Purchase 4.31(1.38) 4.58(1.22) -.27* -2.34* 498 Intentions

Subjective 2.52(1.34) 2.86(1.40) -.33** -2.72** 95 knowledge 498

Note. a Values are based on the arithmetic mean of the raw scale scores; ***p<.001, **p<.01, * p < .05, showing significant difference between groups

Table 14. Ethical Issue Main Effects

Model df df p-value Unconstrained 911.81 430 Full-constrained 935.11 456 23.31 26 .62 Constraint on: H1: Attitude Purchase Intentions 912.48 431 .67 1 .41

H2: PBC Purchase Intentions 912.27 431 .46 1 .50

H3.1: Injunctive norms Attitude 911.81 431 0 1 .99

H3.2: Descriptive norms Attitude 912.60 431 .79 1 .37

H4.1: Injunctive norms Purchase Intentions 912.67 431 .86 1 .35

H4.2: Descriptive norms Purchase Intentions 916.02 431 4.21 1 .04*

H6.1: Objective knowledge Attitude 912.16 431 .35 1 .55

H6.2: Subjective knowledgeAttitude 912.31 431 .50 1 .48

H7.1: Objective knowledge PBC 913.56 431 1.75 1 .19

H7.2: Subjective knowledge PBC 911.88 431 .07 1 .79

Note. * p < .05, indicating significant difference between groups

Table 15. Ethical Issue Moderating Effects

Note. All values are standardized estimates. Significant effect No effect, ** p<.01, *p<.05.

Figure 3. Summary of SEM Results of Study 2

3.8. Discussion

This study examined major determinants of consumers’ attitudes and intentions to purchase socially responsible apparel products based on the TPB (Ajzen, 1991). In particular, the study focused on the influence of different types of social norms and product knowledge on SRCB and tested their independent effects. As in Study 1, the findings from Study 2 confirmed support for the TPB and the inclusion of additional factors, specifically social norms and product knowledge to the TPB model.

First, an interesting finding was that the significant antecedents of attitude and purchase intentions were very similar for organic cotton and fair-trade apparel.

Consumers’ attitudes toward purchasing the two products were identically motivated by injunctive norms, objective knowledge, and subjective knowledge. In other words, when consumers perceive that buying organic cotton or fair-trade apparel is socially approved, when consumers have a higher level of product knowledge, and when they believe that they know more about the product, they are more likely to have positive attitudes towards the purchase. Injunctive norms exerted the strongest influence on attitude among the three factors. Descriptive norms, on the other hand, which involve perceptions of whether significant others typically purchase the products, were not a significant antecedent of attitude formation. Thus, it can be concluded that injunctive norms function as an effective factor in forming positive attitudes toward purchasing socially responsible apparel products, whereas, descriptive norms do not. The results have clearly shown that each type of norm has a different effect on consumers' acceptance of socially responsible apparel products and greater attention needs to be devoted to displaying social approval or social obligations related to SRCB to encourage consumers to form positive attitudes.

Significant antecedents of intentions to purchase organic cotton and fair-trade apparel were also very similar. As the TPB (Ajzen, 1991) predicts, when consumers had more positive attitudes and perceived less barriers to purchasing the products, they were more likely to have higher intentions to purchase organic cotton and fair-trade apparel.

Interestingly, injunctive norms exerted a strong influence on intentions to purchase fair- trade apparel, whereas, it did not have a significant effect on purchase intentions of organic cotton apparel. Descriptive norms, however, strongly influenced purchase intentions of both organic cotton and fair-trade apparel. These results once again provide support for the independent effect of each type of social norm on consumer behavior.

This study demonstrated that inclusion of this additional source of social influence (i.e., descriptive norms) which is not included in the original TPB model helps to better explain consumers purchase intentions of socially responsible products. The results point out the need to take into consideration the social influence that reflects one’s perception of others’ behaviors to increase consumers’ purchase intentions.

The TPB assumes that the relative importance of the constructs differs across behaviors and situations (Ajzen, 1991). Therefore, it is appropriate to identify which TPB predictors have stronger effects on consumers’ intention to purchase in the context of socially responsible apparel products. To begin with, attitude had the strongest effect on purchase intentions of both organic cotton and fair-trade apparel which was consistent with many studies on SRCB (e.g., Han et al., 2010; Kalafatis et al., 1999; Kim &

Karpova, 2010).

Further analysis of other constructs in the TPB model produced contradictory results for the two products. In terms of organic cotton apparel, PBC was a stronger

99 antecedent of purchase intentions than injunctive norms. However, the result was opposite for fair-trade apparel; injunctive norms had a much stronger influence on purchase intentions than PBC. The fair trade result was inconsistent with previous studies that suggested that the influence of subjective norms (i.e., injunctive norms) on intention is generally less significant than PBC (e.g., Sheppard et al., 1988; Van den Putte, 1991).

Considering this particularly strong effect of injunctive norms on purchase intentions, interventions designed to encourage purchase of fair trade apparel need to strongly emphasize the role of social influence.

Furthermore, while injunctive norms had varying results between the two socially responsible products, descriptive norms consistently had very strong effects on purchase intentions; the strength of this effect was much stronger than that of PBC on purchase intentions in both cases. This indicates that when consumers perceive that others are commonly purchasing socially responsible apparel products, it will strongly and positively motivate their own purchase behaviors of such products. The finding was in line with White et al.’s (2009) study that examined the role of descriptive and injunctive norms in household recycling behaviors. They suggested that the two norms are unique predictors of intentions to engage in household recycling and descriptive norms are more powerful in predicting intentions than injunctive norms. These findings indicate that it would be essential to increase the visibility of the purchase of socially responsible products so that consumers can observe others engaging in the target purchase behavior.

Given the results, which are similar to previous studies on focus theory of normative conduct (Cialdini et al., 1991; Reno et al., 1993; Kim, et al., 2012), it can be concluded that descriptive and injunctive norms have independent influences on

100 consumer acceptance of socially responsible apparel products. The strength of the relationship between the two types of social norms was moderately-strong, but SEM results indicated that they uniquely predict attitude and purchase intentions of organic cotton and fair-trade apparel. In sum, injunctive norms seem to be more powerful in encouraging positive attitudes, whereas, descriptive norms function as a stronger motivator to increase intentions to purchase the products. Previous consumer research mostly examined a singular type of norm and that may account for the inconsistency across reported results regarding the influence of social norms on consumer behaviors

(Cialdini et al., 1991; Reno et al., 1993). Therefore, future studies that examine the explanatory value of social norms need to distinguish the concept of descriptive and injunctive norms as each type of social norm reflects a distinct resource of human motivation (Deutsch & Gerard, 1955) that leads to different purchase behaviors.

In addition, independent effects of the two types of product knowledge on attitudes were evidenced by the results which were in line with previous studies that highlighted the need to distinguish knowledge into two types (Selnes & Gronhaug, 1986;

Feick et al., 1992) in consumer studies. Objective and subjective knowledge strongly and independently affected attitude toward purchasing socially responsible apparel products.

Considering the fact that consumers generally have low levels of knowledge about such products (Ha-Brookshire & Norum, 2011; Lin, 2009), the results present practical implications to those who are interested in raising awareness of SRCB.

Although this study did not examine the direct influence of product knowledge on purchase intentions, the importance of consumer education on these products is not negligible as consumers’ intention to purchase is not independently formed; it is strongly

101 influenced by attitudes. Because increasing objective and subjective knowledge are both crucial in forming positive attitudes toward the purchase, finding ways of providing more opportunities to consumers to objectively learn about and to become familiar with socially responsible apparel products would be necessary.

However, in contrast to what was initially expected in the proposed model, the path from product knowledge to PBC was not significant which was inconsistent with findings from Kang et al., (2013) and McEachern and Warnaby (2008). They suggested that when consumers have a higher level of product knowledge, they perceive more control over purchasing the products. Perhaps the reason for this conflicting result may be due to the low availability and high price of organic cotton and fair-trade apparel. For instance, the price of organic cotton apparel is generally 10-30% higher than apparel made of conventional cotton (Goldbach, Seuring, & Back, 2003). Therefore, even though consumers have a considerable amount of knowledge about these products, they could face high barriers to purchase them due to such problems as higher price or lower availability.

This study also found that consumer response to socially responsible apparel products varied based on the type of ethical matter related to them: environmental protection and support for human rights. It seemed that consumers responded more positively to the latter matter than the former: they have higher purchase intentions, believe that others also have higher purchase intentions, and perceive having greater knowledge of fair-trade apparel than organic cotton apparel. This finding was somewhat similar to a previous study (Loreiro & Lotade, 2005) that compared organic and fair-trade

102 coffee and found that the premium consumers are willing to pay was higher for fair-trade coffee than for organic coffee.

Although significant differences were found in the mean scores of the constructs, as explained in the previous paragraph, the multi-group SEM results showed that the study’s causal model did not differ much between the two product types; only one structural path out of a total of ten was reported to differ. The results imply that whether the apparel product is organic or fair-trade seems to be less important during the process of forming attitudes or making purchase decisions.

3.9. Managerial Implications

This study confirmed the potential of supporting socially responsible practices in the apparel industry. It showed that consumers have moderately high intentions to purchase organic cotton and fair-trade apparel in the future; the mean score for purchase intentions was 4.31 and 4.58 on a 7 point scale for organic cotton and fair-trade apparel respectively where higher scores indicate stronger intentions to purchase. Furthermore, while examining the item, ‘willingness to pay more’, was not part of the main analysis, it was included in the questionnaire (Appendix F) to get an overall picture of consumers’ willingness to pay a premium price for organic cotton or fair-trade apparel products because they are generally more expensive than conventional products. The result of willingness to pay more illustrated that even though apparel products with ethical labels sell in a higher price range, demand for them does exist in the current market.

The findings of this study provide important managerial insights to marketers who are interested in promoting socially responsible apparel products by indicating significant determinants of consumers’ purchase behaviors. First of all, marketers should focus on

103 utilizing social norms, since they have a great impact on motivating consumers to purchase the products. For example, this study confirmed that one’s perception of others’ behavior (i.e., descriptive norms) is important when making purchase decisions in the context of ethical consumption. Accordingly, in order to increase SRCB, it would be important to increase the visibility of SRCB. Marketers may consider developing marketing strategies that provide exposure to ethical consumption habits of other people or creating network-based online communities that offer frequent peer interactions related to the target behavior.

Furthermore, studies suggest that combining the normative elements (i.e., injunctive and descriptive norms) would be most effective in promoting ethical behaviors

(e.g., Schultz, Khaziana, & Zaleskia, 2008). The present study confirmed the independent effect of injunctive norms on forming attitude toward the purchase. Accordingly, marketers could combine the implications of injunctive norms to their strategies to corroborate the effect of social norms. For example, they may consider employing normative messages in their marketing campaigns that directly reflect social approval and encourage SRCB.

Also, this study demonstrated that product knowledge significantly affects attitudes towards purchasing socially responsible apparel products. Consistent with previous research (Ha-Brookshire & Norum, 2011; Lin, 2009; Manget et al, 2009), the data from both groups of organic cotton and fair-trade apparel indicated that consumers’ knowledge level regarding these products was considerably low; subjective knowledge scores were 2.52 and 2.86 on a 7 point scale for organic cotton and fair-trade apparel respectively in which higher scores indicate higher subjective knowledge. Considering

104 these results, promotions designed to provide objective information and to increase consumers’ familiarity level and awareness of these products might be an effective way to shape positive attitudes which would, in turn, have a direct effect on purchase intentions.

Finally, this study showed that significant factors that influence consumers’ attitudes and purchase intentions did not differ much between organic cotton and fair- trade apparel. This implies that consumer behaviors toward apparel products that involve environmental protection and support for human rights are closely related. It is possible that consumers simply judge the products by the single issue of ethical consumerism, without precisely distinguishing them into specific ethical matters. Thus, in terms of utilizing social norms or product knowledge to promote SRCB for these products, marketers could apply similar strategies.

3.10. Theoretical Implications

Study 2 focused on examining SRCB in the context of the apparel and textile field, by using the TPB framework to identify significant factors that facilitate the purchase of socially responsible apparel products. This study made theoretical contributions to SRCB research by examining consumers’ purchase behaviors in a more detailed manner. While many previous SRCB studies select one particular type of ethical issue (mostly environmental protection), Study 2 examined two different domains of ethical consumerism (i.e., environmental protection and employment and human rights support) and compared the results. Additionally, Study 2 thoroughly examined the effect of different types of social norms and product knowledge, which have not been well

105 explained in previous apparel studies, and confirmed their independent effects on purchase behaviors of socially responsible apparel products.

3.11. Limitations and Suggestions for Future Research

In spite of the interesting findings and contribution of Study 2, it is important to address its limitations and provide suggestions for future research. First, this study focused deliberately on social norms and product knowledge and it is likely that they are only partial drivers for SRCB. The researcher admits that the model derived from the study’s result may not be complete to fully explain consumers’ purchase intentions of organic cotton or fair-trade apparel. Other constructs, such as product-related attributes and skepticism towards ethical labels may also greatly affect purchase behaviors in this context. In relation to consumer skepticism, a previous study showed that when environmentally friendly products have credible labels, many consumers are willing to pay a premium for the purchase (Hyllegard et al., 2012). While these constructs have been examined in previous studies, their relationships with social norms or product knowledge has not been explained. For example, it is possible that the influence of significant others on making one’s own purchase decision may vary by the level of skepticism related to the purchase. Therefore, future research investigating SRCB, could consider including these additional constructs into the model.

106

Chapter 4: Study 3- Efficacy of Using Social Media to Increase Knowledge and Promote SRCB

4.1. Overview

Many studies on ethical behaviors emphasize the influence of relevant others, such as peers in effectively changing one’s own behavior (e.g., Goldstein et al., 2008;

Kim et al., 2012). These findings imply suggestions for using social media as a marketing tool to motivate SRCB as it amplifies the power of consumer-to-consumer communication by providing an interactive platform through which consumers can easily communicate with many others (Mangold & Faulds, 2009). Numerous companies are using some form of social media websites, however, so far, there is a lack of empirical findings that show how consumers respond to them. One common feature of popular social media websites is that there are various information formats that can be used to present knowledge-building messages to the public. Common representative formats include video, image, and text display. These formats vary by their level of media richness, which involves sensory breadth, or the number of senses that the medium engages (e.g., Coyle & Thorson, 2001; Steuer, 1992). Furthermore, information provided on these websites varies in terms of its quality and quantity. For companies, presenting to consumers the right level of information in terms of its quality and quantity has become more vital because these days, online shoppers expect to find product related information

107 at the click of a button. It would thus be necessary for companies to ensure that appropriate information is easily available to consumers.

4.2. Problem Statement

Because consumers’ familiarity or knowledge level regarding socially responsible products is generally low, identifying information formats and information characteristics that exert a positive impact on consumer knowledge and purchase behaviors will be meaningful to marketers who are interested in promoting their products. Advertising plays a major role in consumer learning (Hoch & Ha, 1986); it may be considered as a means of consumer education given that it is often used to provide information about the product to consumers. Needham, Harper, and Steers (1985) showed that the majority of consumers learned about goods and services through exposure to information in advertising which consequently helped them make better buying decisions. Thus, the major problem Study 3 seeks to examine is the effectiveness of using social media websites which have become an emerging platform for business, mainly focusing on the influence of information formats in terms of their media richness, and information characteristics in terms of quality and quantity of information, as means to increase knowledge about and promote SRCB.

There is currently a vast amount of information offered in the online environment; for companies that adopt online promotion strategies, it would be important to identify what kind of information has a stronger influence on consumers’ decision-making. The results would suggest practical implications to marketers by indicating which online promotion strategies they should focus on to most effectively educate consumers and ultimately motivate consumers to purchase their product.

108

Study 3 combined media richness theory with the hierarchy of effects model to test its relationship with consumers’ cognitive, affective, and conative responses. While the two theories independently have been applied extensively in previous literature, there are very few that integrate both theories to explain media effects on consumer behavior

(e.g., Simon & Peppas; 2004; Xu, Oh, & Teo, 2009). The lack of studies examining the relationship between media richness and consumer behavior and the current popularity of online marketing strategies used by retailers calls for further examination in this area.

Consumer behavior related to ethical consumerism, from awareness to the final stage of the purchase, is not a simple process, so, retailers hoping to promote SRCB using different types of media would need to understand the role that media richness plays in the various stages of consumer response.

4.3. Purpose of Study

The purpose of Study 3 was to examine the effectiveness of using social media websites to increase knowledge about and promote consumers’ socially responsible purchase behavior. Based on the hierarchy of effects model, the researcher identified factors that are useful in motivating purchase behaviors in the social media environment

(Figure 3). A primary focus of this study is placed on the influence of information format and information characteristics. There are different formats for posting available information on social media websites such as video, image, and text, which vary in their level of media richness. This study examined the effect of information format on consumers' cognitive, affective, and conative responses. Will there be evidence that a richer medium (e.g. video posting) is more effective in influencing consumer responses than a leaner medium (e.g. text only posting)? Text only postings and even those that

109 contain 2D images contrast with a rich video posting: Text and 2D images reach consumers through static visual stimuli only, whereas video uses a combination of sight, sound and motion. Consequently, video is acknowledged as the richest medium compared to the other two alternatives in the present study.

Furthermore, in an attempt to identify important determinants of knowledge acquisition related to socially responsible purchases, the effect of information characteristics such as quantity and quality of information was examined. Quantity and quality of information are reported to have significant and independent effects on consumers’ decision making (Keller & Staelin, 1987) which demonstrates the necessity to incorporate these characteristics of information in the model.

4.4. Literature Review

4.4.1. Media Richness

Media richness theory (Daft & Lengel, 1986) is known as a useful framework to explain factors influencing an individual’s decision regarding the choice and usage of a media. According to this theory, media richness is a fixed and objective property of a communication medium and is described as the ability of a medium to reproduce the information that it conveys. Communication media differ in terms of their “richness”; a medium with greater media richness clarifies ambiguous matters and enables people to communicate and promote understanding in a timely and accurate manner. Steuer (1992) explains that two major factors contribute to media richness: the sensory breadth which is the number of sensory dimensions (e.g., sight, sound, touch, smell, and taste) presented and sensory depth which is the quality and resolution within these sensory channels.

110

Hence, messages that appeal to multiple perceptual systems are associated with higher levels of media richness and are generally regarded as more effective than those that involve single sensory systems (Li, Daugherty, & Biocca, 2002). However, in the case where the conveyed information pertains to simple tasks, it is reported that the use of rich media may create conflicting cues, distracting the person’s attention from the information itself (Daft & Lengel, 1986). An overabundance of rich media cues in simple task situations appears to make information processing unnecessarily complex for the information receivers. Therefore, selecting a medium with a level of richness that corresponds to the complexity of the task is an important issue.

In the context of SRCB, it may be useful to examine the role that media richness plays in knowledge acquisition and acceptance of socially responsible products. Carrigan,

Szmigin, and Wright (2004) mention that there is currently an abundance of information in the media related to socially responsible products, however, many consumers have trouble making use of it. It is extremely difficult for consumers to adequately assess the product because marketing efforts from companies have become so complex (Titus &

Bradford, 1996). This may explain the reason consumers state they feel confused when shopping for socially responsible products. A vast majority of consumers tend to rely on the company's media advertisements for product-related information; yet lack of good and reliable information from these sources is likely to be a major barrier to purchasing the products (Manget et al., 2009).

4.4.2. Hierarchy of Effects Model

The hierarchy-of-effects model is predominantly used among various advertising theories as it clearly shows how advertising works. A major advantage of this model is

111 that it identifies factors that are crucial in understanding consumer response. The hierarchy of effects model suggests that human behavior is composed of three dimensions, that is, cognitive, affective, and conative dimensions (Lavidge and Steiner, 1961). The cognitive dimension relates to developing knowledge and awareness, the affective dimension to developing attitudes and feelings, and the conative dimension to developing intention and actual behavior. This model implies that people react to advertising messages in an ordered way: from cognition to affect and next, from affect to conation.

Although Lavidge and Steiner’s (1961) hierarchy model has been frequently cited in previous literature, considerable discrepancy exists considering the order of the three dimensions. Several alternatives to the original model (Lavidge & Steiner, 1961) include

Krugman’s (1965) suggestion of cognition-conation-affect sequence for low involvement situations and Ray et al.’s (1973) conation-affect-cognition sequence, in which purchasing behavior comes first, followed by attitude formation to support their choice, and selective learning to further reinforce their purchase decision. These alternative models suggest that different hierarchical models have been developed to explain diverse conditions of consumer decision making, yet, they all seem to agree on the significance of the three basic dimensions of the hierarchy of effects model (i.e., cognitive, affective, and conative dimensions).

4.4.3. Information Quality and Quantity

The provision of information generally presents significant benefits to consumers in regards to enhanced decision making and satisfaction (Mazis et al. 1981; Zhang &

Fitzsimons, 1999). However, certain aspects of information are necessary for that information to have a positive influence on the process of consumer decision making. A

112 body of research has focused on key characteristics of online information such as information quality and quantity to examine their effects on consumer behavior (e.g.,

Barnes & Vidgen, 2002; Kozup, Howlett, & Pagano, 2008; Keller & Staelin, 1987;

Sivaramakrishnan, Wan & Tang, 2007; Wang, 1998). Consumers’ online shopping decisions rely heavily on information offered on the website because shoppers lack physical contact with the actual product. Hence, quality and quality of online information becomes crucial, most especially in the online shopping environment (McKinney et al.,

2002).

Information quantity can be defined as the amount of information that is available and information quality as the usefulness to consumers in assessing the product (Keller &

Staelin, 1987). High quantity of the information can be characterized as being detailed and complete about a specific matter (Kozup, et al., 2008; Sivaramakrishnan et al., 2007).

On the other hand, high quality of information can be achieved when the information content is accurate, timely, believable, relevant, and easy to understand (e.g., Barnes &

Vidgen, 2002; Wang, 1998).

The critical role of information regarding ethical issues in SRCB has been discussed in many studies. For instance, Yiridoe, Bonti-Ankomah, and Martin (2005) indicated that consumers are highly dependent on information provided by the manufacturers or retailers because in general, they cannot detect the presence of the product’s ethical characteristics even after purchasing and while using the product. De

Pelsmacker and Janssens (2007) mentioned that products associated with ethical claims often suffer from a low level of credibility due to a lack of information available to consumers, in terms of its quality and quantity. Furthermore, Shaw and Shiu (2002)

113 suggested that future studies can make improvements in understanding SRCB by exploring the impact of information on buying behaviors.

4.4.4. Objective and Subjective Knowledge

The importance of product knowledge in consumer behavior has been well supported in previous literature (e.g., Brucks, 1985; Sujan, 1985). These studies show that product knowledge influences various phases of consumer response behaviors such as product evaluation, satisfaction, and purchase intention. Park and Lessig (1981) suggested two major approaches regarding the measurement of knowledge: (a) measuring how much an individual actually knows about the product (i.e., objective knowledge) and

(b) measuring how much an individual thinks he/she knows about the product (i.e., subjective knowledge). Similarly, Brucks (1985) illustrated three categories of product knowledge that are employed in consumer behavior research: (a) objective knowledge, (b) subjective knowledge, and (c) prior experience with the product. However, Brucks (1985) explained that experience-based measures of this construct are less directly associated with behaviors.

The measurement of knowledge has long been a subject in consumer behavior literature and a majority of studies that examine the effect of knowledge often treat it as a single construct, generally measuring one of the two knowledge types: objective or subjective knowledge. Because each knowledge type is known to influence consumers in a different way (Selnes & Gronhaug, 1986; Feick et al., 1992), the distinction between the types of knowledge would be important in understanding consumer behavior. In spite of the growing body of literature concerning SRCB, there is a lack of studies that specifically examine the difference between the impact of objective and subjective

114 knowledge on consumers’ acceptance of socially responsible products. Therefore, it is necessary to differentiate and examine the impact of the two types of knowledge in this context.

Figure 4. Conceptual Model of Study 3 (Hierarchy of Effects Model)

4.5. Conceptual Model and Hypotheses

4.5.1. Effect of Media Richness on Consumer Response

There are several ways companies can post contents in the social media environment; the most widely used formats tend to be text, image, and video. These formats vary in terms of their level of media richness. According to previous studies (e.g.,

Coyle & Thorson, 2001; Steuer, 1992), media richness can be operationalized by sensory breadth, the number of senses that the medium engages. Because the video, using the combination of sight, sound and motion, reaches the viewer with the highest number of 115 senses, it could be considered as the richest medium compared to other formats such as text or photo postings.

Exploratory findings from Cvijikj, Spiegler, and Michahelles’s (2011) study suggest that there is a significant effect of information format on consumer behavior.

Major differences were found across information formats regarding how consumers react on Facebook. For example, the number of likes, number of comments and interaction duration turned out to be much higher for a video posting (rich medium) than a photo posting (lean medium).

Previously, sales representatives played a major role in providing product information to customers (Kim & Stoel, 2005), whereas now their role in consumer education may have been reduced as people have easy access to product and service information provided on retail websites and real-time feedback from other customers in various social media outlets. It is reported that an online store with a higher degree of media richness can positively affect consumers’ cognitive response. Suh and Lee (2005) found that a rich medium (i.e., virtual reality interface with high-quality 3-D product images) provides a more effective learning environment about the product to the viewers than a lean medium (i.e., 2-D static images). Similarly, a large body of research applied media richness theory to explain the effectiveness of various media types in remote learning systems outside the traditional classroom environment. For example, studies found that higher levels of media richness lead to more satisfaction (Valacich, Mennecke,

Wachter, & Wheeler, 1994), more communication and interaction with others (Shepherd,

& Martz, 2006), and stronger intentions to use the medium (Lai & Chang, 2011).

116

Additionally, previous studies stated that varying levels of media richness produce different results in consumers’ affective response. Li, Daugherty, and Biocca

(2003) found that presenting a product using 3-dimensional images assists in creating positive emotion for consumers in the online environment. Similarly, Park et al. (2008) created two conditions of online product presentation (rotating vs. non-rotating) and found evidence that rotation influences various types of consumer response which included emotion as well as information quantity perceptions and purchase intentions.

H1: Consumers’ cognitive, affective, and conative responses and perceptions of information quantity and quality will differ based on information format (i.e., media richness).That is, the means of constructs will vary across information formats.

4.5.2. Effect of Information Quantity and Quality on Knowledge

Consumers commonly make judgments about a purchase based on the product related information they acquire from manufacturers or retailers (Yiridoe et al., 2005), and such information varies in terms of quantity and quality. Information quantity may be defined as the amount of information that is available and information quality is defined as the usefulness to consumers in assessing the product (Keller & Staelin, 1987).

Previous research has demonstrated that information quantity and quality affect consumer responses toward the purchase. Zhang and Fitzsimons (1999) found that when consumers were presented with a large amount of information, their satisfaction level increased. Keller and Staelin's (1987) study suggested that information quality significantly influences consumers’ cognitive ability to make more effective purchase decisions. Although many studies in the field of consumer behavior have been directed toward investigating the effect of information, extant literature lacks empirical findings 117 that determine the direct effect of information quantity and quality on consumers’ knowledge. To address this gap, this study posited that perceptions of information quantity and quality evoked by the contents of social media will influence consumers’ objective and subjective knowledge level.

H2: Perception of information characteristics will have a positive and direct effect on knowledge.

H2.1: The level of information quantity perceived will have a positive and direct effect on objective knowledge.

H2.2: The level of information quantity perceived will have a positive and direct effect on subjective knowledge.

H2.3: The level of information quality perceived will have a positive and direct effect on objective knowledge.

H2.4: The level of information quality perceived will have a positive and direct effect on subjective knowledge.

4.5.3. Relationships of Knowledge, Emotion, and Purchase Intention

The hierarchy of effects model has obtained great attention from both academic scholars and practitioners (Barry & Howard 1990). Since the model works as a conceptual tool to predict consumer behavior, it forms a good basis for advertising planning by presenting information about where advertising strategies need to focus. The hierarchy of effects model (Lavidge & Steiner, 1961) suggests that consumers respond to promotional messages in an ordered way, starting from unawareness of the product to the actual purchase. This model hypothesizes a causal relationship from cognitive stage (e.g.,

118 understanding and learning) to affective stage (e.g., feeling), and from affective stage to conative stage (e.g., taking action).

Most promotional messages associate the product with positively valued traits. As consumers learn about such associations, the messages come to be represented in their memory as product-related beliefs. The more the product is linked to positive traits, the more favorably disposed consumers will be toward the product purchase (Smith, Chen, &

Yang, 2008). In addition, the presence of favorable product attitudes results in stronger purchase intentions because one’s affective states directly impact product evaluations

(Peracchio & Meyers-Levy, 1994). Yang and Smith (2009) found that consumers’ positive emotion about an advertisement directly influence purchase intentions of the advertised product.

Similar results have been found in studies examining the effect of store atmospherics on consumers’ cognitive, affective, and conative responses. Fiore, Yah, and

Yoh (2000) found that in-store product displays using music or fragrance evoke a positive affective state and, in turn, result in a stronger purchase intention. Similarly,

Spies, Hesse, and Loesch (1997) found that consumers who feel pleasant during their shopping experience have stronger purchase intentions than those who feel unpleasant. In addition, studies on online store atmospherics have suggested that using advanced tools such as interactive functions or 3D product presentation provide better or more information, and influence consumers’ emotion and purchase decisions (Park, Stoel, &

Lennon, 2008).

Although there are indications in previous literature (e.g., Lavidge and Steiner,

1961) that development of knowledge (i.e., cognitive dimension) influences forming

119 positive emotion (i.e., affective dimension), they generally do not distinguish between objective and subjective knowledge to explain their independent effects. The distinction between the two types of knowledge may be important in understanding SRCB as each type influences consumers differently (Selnes & Gronhaug, 1986; Feick et al., 1992).

H3: Knowledge will have a positive and direct effect on emotion.

H3.1: Objective knowledge will have a positive and direct effect on emotion.

H3.2: Subjective knowledge will have a positive and direct effect on emotion.

H4: Emotion will have a positive and direct effect on purchase intentions.

4.6. Methods

4.6.1. Instrument

An online questionnaire was developed to measure each construct included in the present study. All questions were rated on a 7 point Likert scale (1=strongly disagree,

7=strongly agree) excluding true/false response questions that measured objective knowledge. The complete list of the items for measuring the constructs is provided in

Appendix C.

 Information quality and quantity- Five items measured perceptions of information quality; they were adapted from Barnes and Vidgen’s (2002) study. The scale included statements like “Information contained on the posting provides accurate information” and respondents were asked to indicate the level of agreement on a seven-point scale.

Reliability of the scale was assessed by Cronbach’s alpha coefficient which was reported to be .89 in this study, showing a satisfactory level of consistency among the items. To

120 measure perceptions of information quantity, 1 item from Sivaramakrishnan et al. (2007) and 2 items from Kozup et al.’s (2008) study were used. The scale included statements like “In my opinion, the information about fair-trade provided by the posting is…” and were rated on a 7 point scale (1=very brief, 7=very detailed). Two items that measured information quantity in Kozup et al.’s (2008) study showed sufficient level of reliability; the correlation coefficient between the two items was .78. Sivaramakrishnan et al. (2007) used a single item to measure this construct.

Trade International, 2013; Fair Trade USA, 2013). Similar to Aertsens et al (2011)’s measure of objective knowledge, respondents indicated how certain they are of each of their true/ false response on a 5 point scale (1= uncertain, 5= certain) to take into account participant guessing. Respondents who are guessing or, in other words, less certain of the answer received lower scores than those who are more certain of the answer. Following

Aertsens et al (2011)’s suggestion, the objective knowledge score was calculated as follows: “ a wrong answer with a certainty of 5, resulted in a score of 0; a wrong answer with a certainty of 4, in a score of 1; a wrong answer with a certainty of 3 in a score of 2, and so on; a correct answer with a certainty of 1 resulted in a score of 5; a correct answer with a certainty of 2, in a score of 6, and so on. The maximum score is thus given to a correct answer with a certainty of 5, which results in a score of 9. The total objective knowledge score is then calculated by summing the scores on each of the four statements and these therefore range between 0 and 36” (p.1360). Aertsens et al (2011) calculated 121 the Cronbach’s alpha score which was .61, indicating a sufficient level of consistency among the four items.

To measure participant’s subjective knowledge, a scale developed by Flynn and

Goldsmith (1999) was used. It included statements like “When it comes to _____, I really don’t know a lot”. Reliability of the scale in their study was assessed by Cronbach’s alpha across three product domains (i.e., clothing, movies, and music) which ranged from .88 to .94.

 Emotion- 6 items from Emotional Quotient (EQ) scale adapted from Wells’s (1964) study were used to measure respondent’s affective response. It included statements like “This posting is very appealing to me.” and respondents were asked to indicate the level of agreement using 7 point scale. While Well’s (1964) original study did not report Cronbach's alpha coefficient of the Emotional Quotient (EQ) scale, many subsequent studies that use this scale have supported a sufficient level of reliability. For instance, in Kim, Forney, and Arnold’s (1997) study, Cronbach’s (1951) alpha coefficients for measuring emotion using Emotional Quotient (EQ) scale ranged from .90 to .96.

 Purchase intention- To assess purchase intentions, the respondents indicated on a seven-point scale the extent to which they intend to purchase the fair-trade product. It included items like “I would like to purchase _____ in the future.” (Conner et al., 1999;

Kang et al., 2013). Cronbach's alpha for the intention scale was .75 in Conner et al.’s

(1999) study and .87 in Kang et al.’s (2013) study.

4.6.2. Procedures

To test the effect of media richness, three distinct formats of information were created (i.e., video, image and text combined, and text only). First, a video that was made

122 by a reliable source (i.e., Fair trade USA; http://www.youtube.com/watch?v=_z3PkCnEsno) was used for the study. This video provided some general information about the meaning of fair trade, working conditions of the employees and how fair trade products are produced. Based on the video the other two formats were created. Image and text combined format (Appendix D) had four images captured from the video and text directly obtained by transcribing the narration of the video. Text only format (Appendix E) used the same text as the image and text combined format. Prior to the experiment, three judges (graduate students in consumer studies) assessed the contents of the three formats to ensure they all contained equal contents. Because the text from the latter two formats was directly taken from the narration of the video format, the agreement regarding equal contents across the three formats was easily achievable. After carefully reviewing the three formats that were created for the experiment, the judges perfectly agreed that they contained identical contents. Each participant was exposed to one of the three information formats for a limited amount of time (duration of the exposure time was controlled to be the same for all information formats). After viewing the posting about fair-trade, they were asked to complete the questionnaire.

Auger and Devinney (2007) state that social desirability bias is inherent in self- reported questionnaires of ethical consumerism. That is, when individuals are faced with answering questions on sensitive topics such as their socially responsible purchase behaviors, they tend to provide answers in a way that will be viewed more favorably by others. Study 3 used a computer-based online method; there is less interaction with the interviewers, and thus, less desire to impress the interviewers. While this study remains

123 susceptible to some level of social desirability bias, the online questionnaire was developed to reduce this effect.

 Manipulation check variable- media richness In this study, media richness is operationalized by sensory breadth, the number of senses which a medium engages, similar to previous studies (e.g., Coyle & Thorson, 2001;

Steuer, 1992). To see whether different levels of media richness were manipulated properly, participants were asked to indicate their agreement/disagreement with the following statements: 1. The posting included images; 2. The posting included moving images; 3. The posting included sounds. These statements were presented to the participants immediately after they had viewed the posting. First, nearly all the participants who were assigned to the video posting indicated that the posting included moving images (97.9%) and sounds (99.3%). Next, 98.1% of the participants who were assigned to the text and image combined posting agreed that the posting included images,

92.9% agreed that there were no moving images, and 93.5% agreed that the posting did not contain any sounds. Finally, all participants who were assigned to the text only posting indicated that the posting did not include any moving images or sounds (100%);

99.4% of them agreed that it did not include any images. These results suggested that the manipulation of media richness was successful in that participants perceived video as the richest medium which engaged the highest number of sensory dimensions followed by image and text combined, and text only.

Pre-test

At the beginning of the experiment, participants took a pre-test which included items that measured objective knowledge and pre-existing attitude regarding fair trade 124 products. The purpose of this pre-test was to see how much the participant already knows and feels about the topic that may potentially have an impact on the study results. The scores were checked to examine whether significant differences existed among the three conditions of information format. One-way between subjects analysis of variance

(ANOVA) was conducted in order to ensure that the three groups did not differ significantly in terms of their objective knowledge level and attitude before the exposure to the posting. The results indicated that the groups did not differ regarding the two variables: objective knowledge, F(2, 452)= .08, p=.93 and attitude, F(2, 452)= .58, p=.56.

In addition, demographic characteristics were compared to further test the homogeneity of the groups. The following demographic variables did not differ significantly across groups: age, F(2, 452)= .26, p=.78; income, F(2, 452)= 1.66, p= .19; education level, F(2, 452)= .08, p= .92; gender, (2, N = 455) = .46, p = .80; race,

(10, N = 455) = 10.69, p = .38.

4.6.3. Sample

A total of 455 participants from the Generation Y cohort were obtained from the

MTurk to test the hypotheses of Study 3. This sample size satisfied the recommended minimum number needed for structural equation modeling (SEM) analysis: n 50 -

450r + 1100 (r= p/ k when p is the number of indicators and k is the number of latent variables, r is 4.6 in this study) (Boomsma, 1982; Marsh, Hau, Balla, & Grayson, 1988).

The demographic information of the Study 3 sample is summarized in Table 16.

There were approximately equal numbers of male (50.1%) and female (49.9%) participants. A majority of participants were between 25 and 34 years of age (63.7%) and

Caucasian (74.1%) with personal income of less than $35,000 (66.1%). Nearly one third

125 of the participants (33.0%) had some college education and a slightly higher number of participants (38%) had obtained a bachelor’s degree.

Although not directly comparable with the sample from the current study which only involved Generation Y participants, recent census data (US Census Bureau, 2013) illustrated that 49.2% of the U.S. population were male, 77.7% were white, 28.5% obtained a bachelor’s degree or higher, and the median personal income was approximately $28,000. The sample from this study and the U.S. population had similar demographic characteristics with the exception of the education level. The sample seemed to have higher education level than the U.S. population; almost half (48.1%) of the sample had obtained a bachelor’s degree or a higher degree which was considerably different from the 28.5% from the U.S. census data.

. A comprehensive survey about Generation Y by Pew Research Center (2010) showed that 61% of the Generation Y people were non-Hispanic whites and 54% had at least some college education. The survey indicated that Generation Y cohort is more ethnically and racially diverse than previous generation cohorts. Regarding the Baby

Boomers (born from1946 to1964), shares of the population for the black and Hispanic were11% and 10% respectively; whereas, they were 13% and 19% respectively among

Generation Y cohort. Compared with this demographic information of Generation Y, the sample of this study had more white participants with higher educational attainment

126

Variable Frequency % Gender Male 228 50.1 Female 227 49.9 Age 20-24 92 20.2 25-29 163 35.8 30-34 127 27.9 35-37 73 16.0 Race White 337 74.1 Black or African American 37 8.1 Hispanic or Latino 33 7.3 American Indian or Alaska Native 2 .4 Asian 43 9.5 Other 3 .7 Annual personal income Less than $10,000 98 21.5 $10,000 - $24,999 123 27.0 $25,000 - $34,999 80 17.6 $35,000 - $49,999 71 15.6 $50,000 to $74,999 63 13.8 More than $75,000 20 4.4 Education Did Not Complete High School 1 .2 High School/GED 47 10.3 Some College 150 33.0 Associate Degree 38 8.4 Bachelor's Degree 175 38.5 Master's Degree 37 8.1 Advanced Graduate/Professional work 7 1.5 or Ph.D.

Table 16. Demographic Summary Statistics of the Sample (n=455)

127

4.6.4. Data Analysis

All statistical analyses for Study 3 were performed using SPSS 19 and AMOS 21.

One-way between subjects ANOVA was conducted to compare the mean and standard deviation of each construct across groups. Post-hoc Tukey tests further determined how each group differed from other groups. Furthermore, in order to test the cause-effect relationships of the constructs, SEM was conducted. The two-step approach, a measurement and a structural model, recommended by Anderson and Gerbing (1988) was adopted.

4.7. Results

4.7.1. ANOVA

One-way between subjects ANOVA was conducted for testing H1, which proposed that consumers’ cognitive, affective, and conative responses will differ based on the information format. The mean scores of constructs across information formats were compared in the analysis. As Table 17 illustrates, the results showed that the effect of information format on perceptions of information quality, F(2, 452)= 3.16 p= .04, subjective knowledge, F(2, 452)= 3.17, p= .04, and emotion, F(2, 452)= 5.49, p= .004, were significant. Post hoc Tukey tests revealed that the participants in the video group

(information quality: M= 5.92, SD= .78; subjective knowledge M= 3.45, SD= 1.34; emotion: M= 5.47, SD: 1.16) scored significantly higher than those in the text only group

(information quality: M= 5.67, SD: .83; subjective knowledge: M= 3.10, SD= 1.30; emotion: M= 5.00, SD: 1.37) for the three constructs. There were no significant differences across information formats regarding information quantity, objective knowledge, and purchase intention scores. Thus, H1 was partially supported.

128

Videoa Image+Texta Texta Constructs F p-level (N= 143) (N= 155) (N= 157) Information 5.92 5.81 5.67 3.16* .04* quality (.78) (.91) (.83)

Information 5.11 4.86 4.93 1.67 .19 quantity (1.19) (1.23) (1.21)

Objective 27.97 28.21 26.99 2.09 .13 knowledge (5.30) (5.77) (5.59)

Subjective 3.45 3.39 3.10 3.17* .04* knowledge (1.34) (1.34) (1.30)

5.47 5.25 5.00 Emotion 5.49** .004** (1.16) (1.21) (1.37)

Purchase 5.61 5.59 5.49 .41 .66 intentions (1.13) (1.17) (1.32) Note. a Values based on the mean of the raw scale scores; Standard deviations in parentheses. **p<.01, * p < .05, showing significant difference between groups.

Table 17. Media Richness Main Effects (ANOVA)

4.7.2. Measurement Model

Confirmatory factor analyses (CFA) were conducted on the 21 indicators of the 5 latent constructs (Figure 5). Information quality, information quantity, subjective knowledge, emotion, and purchase intentions were latent constructs that were included for the CFA. Objective knowledge, on the other hand, was a composite score as described in the methods section. Therefore, it was excluded from the CFA.

This study applied Hair, Black, Babin, and Anderson’s (2010) criteria for determining reasonable fit: /df <3, CFI>.90, TLI>.90 and RMSEA<.08. The fit of the measurement model was within the thresholds of reasonable fit ( /df = 2.86, p<.01,

CFI= .96, TLI= .95, RMSEA= .06). 129

Table 18 reports the reliabilities and the average variance extracted (AVE) values of constructs in the measurement model of Study 3. Cronbach's alpha and composite reliability both ranged from .89 to .95 displaying satisfactory levels of reliability. All constructs’ AVE values exceeded the .50 threshold which was recommended by Fornell and Larcker (1981). The standardized factor loadings were significant and exceeded the .60 threshold, confirming strong convergent validity (Anderson & Gerbing, 1988).

Furthermore, discriminant validity (Fornell & Larcker, 1981) was confirmed by examining the AVE values which were larger than the squared correlations between the constructs, as shown in Table 19.

130

Figure 5. Confirmatory Factor Analysis Model of Study 3

131

4.7.3. Structural Model

The relationships among the latent constructs are assessed to examine the effect of information on knowledge and causal relationships of the constructs based on the hierarchy of effects model (knowledge  emotion  purchase intention); this analysis tests hypotheses H2-H4.The statistical results of the SEM test using the proposed model

(Figure 4) achieved a mediocre fit: /df= 3.75, p<.01, CFI= .93, TLI= .92, RMSEA

= .08. Thus, the modification indices were examined to determine if there were ways to improve the fit of the model. The indices suggested two additional relationships: a path from information quality to emotion and a path from information quantity to emotion

(Figure 6). These paths indicated that participants’ perceptions about the quality and quantity of the information had direct influences on increasing positive emotions.

Previous studies have illustrated that emotion has a direct relationship with the store environment (e.g., Bitner, 1992) and media richness (e.g., Jeandrain, 2001). In addition,

Park et al. (2008) found that one’s perception about information quantity directly influences forming positive or negative emotions when shopping online. Based on these previous findings and conceptual considerations, the initial model was revised to include the two additional paths (i.e., information quality  emotion, information quantity  emotion). After the revision, goodness of fit indices indicated that the fit of the model was within the satisfactory threshold; /df= 2.84, p<.01, CFI= .96, TLI= .95, RMSEA

= .06.

132

Factor α CR AVE loadings Information .89 .89 .61 quality Information contained on the posting…

provides accurate information .81

provides believable information .77

provides current and timely information .79

provides relevant information .83

presents the information in an appropriate .70

133 format

Information .89 .90 .74 quantity In my opinion, the information about fair- .85 trade provided by the posting is (very brief- very detailed).

The amount of information about fair-trade .89 in the posting was (a little-a lot). Continued

Table 18. Result of Measurement Model

133

Table 18 continued

The information about fair-trade provided .84 by the posting seems (not at all complete- very complete). Subjective .90 .90 .70 knowledge I know a lot about fair trade. .93

I feel knowledgeable about fair trade. .89

Among my circle of friends, I’m one of the .75 “experts” on fair trade.

134 Compared to most other people, I know .77 more about fair trade.

Emotion .95 .95 .75 This posting is very appealing to me. .85

This is a heart-warming posting. .83

I like this posting. .91

This posting makes me feel good. .87

This is a wonderful posting. .90 Continued

134

Table 18 continued

This is a fascinating posting. .82 Purchase .93 .94 .83 Intentions I would like to purchase fair trade apparel .91 in the future.

If I see organic fair trade apparel, I intend to .94 purchase or consider purchasing it.

If I see a retail store selling fair trade .89 apparel, I intend to visit the store to purchase a product.

135

1 2 3 4 5 6 1. Information quality .61 .28 .08 .06 .41 .31 2. Information quantity .53 .74 .01 .12 .30 .14 3. Objective knowledge .29 .12 - .00 .08 .08 4. Subjective knowledge .24 .35 .07 .70 .10 .07 5. Emotion .64 .55 .29 .32 .75 .61 6. Purchase Intentions .56 .38 .29 .27 .78 .83 Note. Diagonal elements are AVEs, below the diagonal are correlation estimates, and above the diagonal are squared correlations. AVE of objective knowledge not recorded because it is a composite score.

Table 19. Measurement model: Correlations between Latent Variables

4.7.4. Hypotheses Tests (H2-H4)

When examining the relationship between information and knowledge, the results

(Figure 6) showed that information quantity had a significant effect on subjective knowledge (β=.31, p<.001). However, its effect on objective knowledge turned out to be insignificant (β=-.02, p=.72). Information quality, on the other hand, had a significant effect on both subjective knowledge (β=.10, p<.001) and objective knowledge (β=.30, p<.001). As a result, H2.2 (information quantity  subjective knowledge), H2.3

(information quality  objective knowledge), and H2.4 (information quality  subjective knowledge) were supported, whereas H2.1 (information quantity  objective knowledge) was not.

As hypothesized based on the hierarchy-of-effects model, objective knowledge

(β=.15, p<.001) and subjective knowledge (β=.12, p=006) had strong influences on emotion and emotion, in turn, had a strong influence on purchase intentions of fair-trade products (β=.77, p<.001). In conclusion, the statistical results from the structural model

136 showed that H3.1 (objective knowledge  emotion), H3.2 (subjective knowledge  emotion) and H4 (emotion  purchase intentions) were supported (Table 20).

The two paths that were added after the revision were also significant. That is, information quantity (β=.30, p<.001) and quality (β=.38, p<.001) had direct and strong effects on emotion.

Furthermore, decomposition tests using the bootstrapping method (with 2,000 bootstrap samples) were conducted to examine the mediation effects. A summary of total, direct and indirect effects are presented in Table 21. First, there was a significant indirect effect of information quantity on emotion (standardized estimate = .03) and information quality on emotion (standardized estimate = .06) through knowledge. The bias corrected

95% confidence intervals for these indirect effects excluded the value of zero (.01 - .08 and .03 - .09 respectively), showing that knowledge significantly mediates the relationship between information and emotion. The bias corrected 95% confidence interval provides a range of the true indirect effects in the population and containing zero within this interval range indicates that the true indirect effect is zero (no mediation). In addition, the results showed a significant indirect effect from information quantity to purchase intentions (standardized estimate = .25; 95% confidence interval .16 - .35) and from information quality to purchase intentions (standardized estimate = .39; 95% confidence interval .30 - .50), with knowledge and emotion being the mediators.

137

Information

Quantity .30**

-.02

Objective .31** .15** Knowledge Purchase Emotion Intentions .77** Subjective .30** .12** 138 Knowledge

.10*

.48** Information

Quality

Cognitive Response Affective Response Conative Response

Response Note. All values are standardized estimates. Significant effect No effect, ** p<.01, * p<.05.

Figure 6. Summary of SEM Results of Study 3

138

Standardized Hypothesized path path t-value Results coefficients H2.1: Information quantity  Objective knowledge -.02 -.35 not supported

H2.2: Information quantity  Subjective knowledge .31** 6.24 supported

H2.3: Information quality  Objective knowledge .30** 6.12 supported

H2.4: Information quality  Subjective knowledge .10* 2.04 supported

H3.1: Objective knowledge  Emotion .15** 3.72 supported

H3.2: Subjective knowledge  Emotion .12** 2.73 supported

 H4: Emotion Purchase Intentions .77** 17.02 supported

Information quantity  Emotion .30** 6.70

Information quality Emotion .48** 9.98 Note. **p<.01, *p<.05.

Table 20. Hypotheses Testing Results

139

Independent variables Dependent variables Total Direct Indirect effects effects effects Information Quantity Objective Knowledge -.02 -.02

Subjective Knowledge .31 .31

Emotion .34 .31 .03 (95% CI: .01-.08)

Purchase Intentions .16 -.09 .25 (95% CI: .16-.35)

Information Quality Objective Knowledge .30 .30

Subjective Knowledge .10 .10

Emotion .53 .47 .06 (95% CI: .03-.09)

Purchase Intentions .50 .11 .39 (95% CI: .30-.50)

Objective Knowledge Emotion .15 .15

Purchase Intentions .15 .04 .11 (95% CI: .04-.17)

Subjective Emotion .11 .12 Knowledge Purchase Intentions .11 .03 .08 (95% CI: .02-.15)

Emotion Purchase Intentions .74 .74

Table 21. Standardized Total Effects, Direct Effects, and Indirect Effects

140

4.8. Discussion

Study 3 examined the effectiveness of using social media websites as a way to motivate purchase behaviors of fair-trade products, in particular, focusing on the influence of information format and characteristics. Based on the media richness theory and the hierarchy of effects model, this study tested how information in the media can affect the viewers’ cognitive, affective, and conative responses.

First of all, ANOVA results showed that information format has a main effect on perceptions of information quality, beliefs about subjective knowledge and emotion. In other words, perception about the quality of the information contained in the posting, self-rated knowledge about fair-trade, and emotion toward the posting varied across information formats. Participants who were exposed to the video format (rich media) had significantly higher scores on these three constructs than those exposed to the text only format (lean media). This was consistent with previous studies on media richness (e.g.,

Jeandrain, 2001; Park et al, 2008) that suggested rich media has a stronger influence on consumer response than lean media. Therefore, it would be reasonable to conclude that presenting the information in a format that involves more sensory dimensions (e.g., sight and sound), such as utilizing video, would be effective in providing a learning environment to consumers and to positively motivate them to purchase fair-trade products in the social media website context.

In contrast, ANOVA results showed that information format did not have a main effect on perceptions of information quantity, objective knowledge, or purchase intention.

Identical contents and equal exposure time across the three formats may have led the 141 participants to perceive the quantity of the information to be the same and to have similar objective knowledge test scores. Also, the insignificant effect of media richness on purchase intention illustrates that information format itself does not have a direct influence on consumers’ purchase decision; this finding calls for further investigation to clarify the purchase process for fair-trade products and identify important determinants of purchase intentions, some of which was addressed in the subsequent SEM analysis.

Next, this study demonstrated that the hierarchy of effects model provides a good framework for explaining consumer response toward fair-trade products. The SEM results clearly identified factors that are important in understanding SRCB related to fair- trade products and illustrated that consumers react to information in an ordered way: cognitive dimension  affective dimension  conative dimension. These paths demonstrate that receiving knowledge about fair-trade products evokes positive emotions and in turn results in stronger purchase intentions; these results were similar to findings from previous studies that supported the structure of consumers’ response system based on the hierarchy of effects model (Fiore et al., 2000; Park et al., 2008; Spies et al., 1997).

The original hierarchy of effects model does not include antecedents of the cognitive dimension. Since developing knowledge and awareness is presumed to play a key role in SRCB, this study added characteristics of information to the model to examine their influence on one's objective and subjective knowledge level. The results showed that while the quantity of the information affected only subjective knowledge, the quality of information affected both objective and subjective knowledge. Because participants were exposed to postings containing equal amounts of information about

142 fair-trade regardless of which format they were assigned to, information quantity may have had less impact on their objective knowledge scores. However, as ANOVA results shows, perceptions of information quality varied significantly across information formats and served as a strong determinant of objective and subjective knowledge. The result is somewhat in line with Keller and Staelin's (1987) study that suggested information quantity and quality have independent effects on individual’s decision making. These findings may be helpful to those that hope to promote ethical consumerism considering the vast amount of information offered related to ethical products. Identifying types of information that have a stronger influence on consumers’ knowledge level associated with the purchase and motivating purchase behaviors may be important for developing online promotion strategies.

An important finding of this study is that it confirmed the strong influence of information on consumer behavior. All hypothesized paths that were related to testing the effect of information were significant except information quantity to objective knowledge.

First, perceptions about quantity of the information had a strong impact on one’s subjective knowledge level. This means that when consumers perceive that they are exposed to a quantity of information about fair-trade, they tend to feel more knowledgeable about the topic. On the other hand, perceptions about quality of the information significantly impacted both objective and subjective knowledge. That is, when consumers believe that the information contained in the posting is detailed and useful, they are more likely to obtain more knowledge and also perceive themselves as being more knowledgeable about fair-trade products. Therefore, these results confirm the

143 crucial role of quantity and quality of information in improving one’s knowledge level related to fair-trade products. The results from this study clearly showed that consumers know little about the concept of fair-trade. Similar to Study 2 and previous studies on socially responsible products (e.g., Ha-Brookshire & Norum, 2011; Manget et al, 2009), knowledge levels were considerably low among the study’s participants; the average score of subjective knowledge was 3.31 on a 7 point scale in which higher scores mean higher knowledge. Taking into account the problem of low knowledge about or familiarity with fair-trade concepts, improving the content of the postings on fair-trade websites in terms of quantity and most especially, quality would be meaningful.

This study further found that both information quantity and quality also had a very strong influence on forming emotion, which was consistent with Park et al.’s (2008) findings. Higher perceptions of information quantity and quality led to more positive emotions toward the posting; respondents thought the posting was, for example, more appealing and heart-warming, and believed it made them feel better. Consistent with the suggestions from the hierarchy of effects model, consumers’ knowledge, both objective and subjective, significantly affected emotion. However, as illustrated in Figure 6, the strength of the effect on emotion was much more powerful for information: while the path coefficient to emotion was .15 and .12 for objective and subjective knowledge respectively, it was .30 and .48 for information quantity and quality respectively. Park et al. (2008) suggested that providing consumers with elements to enhance their emotional states in the online shopping environment is a good way to formulate purchase intentions.

Given the direct and strong impact of emotion on purchase intentions, also shown in this

144 study’s hypothesis testing, these results once again highlight the importance of improving the quantity and quality of information in the social media environment. Positive emotion can be induced in many ways, and results of the current study directly suggest that emotion can be strongly enhanced by improving quantity and quality of information of the website contents.

4.9. Managerial Implications

Szymanski and Hise (2000) mentioned that consumers use online websites not only to purchase but also to obtain product related information. Consumers rely heavily on information and images displayed on online websites in all phases of the purchase decision process. Thus, identifying formats and characteristics of information that are most useful for enhancing knowledge about the product and motivating people to purchase will assist marketers in better serving their customers and developing marketing strategies.

The findings of Study 3 suggest that both quantity and quality of information play a significant role in consumers’ acceptance of fair-trade products. They had positive relationships with both consumer knowledge and emotion, which in turn had a great influence on intentions to purchase fair-trade products. In the current study, the concept of information quality represents accurate and timely information about fair-trade that is conveyed via an online posting. Information quantity refers to the amount of information about fair-trade. Large quantity does not necessarily mean lengthy text or long viewing time of the video; it rather involves detailed and complete information about the fair- trade concepts. Based on these findings, fair-trade retailers should focus on improving the

145 information content, in terms of its quality and quantity, of their social media websites to increase consumer knowledge about the fair-trade concept and thus, motivate sales.

This study also found that emotion toward the posting had a very strong effect on purchase intentions of fair-trade products. In other words, when consumers formed positive emotions about the posting related to fair-trade, they were more likely to have stronger intentions to purchase fair-trade products. Because fair-trade products are associated with ethical issues such as support for human rights and fair employment, it is plausible that consumers rely heavily on their affective responses when making purchase decisions. This study found that a rich medium that include moving images and sounds is more effective than a lean medium that simply consisted of text in forming positive emotions. As a result, fair-trade marketers may consider applying richer media (e.g., video) in their online websites to positively motivate consumers’ emotion regarding fair- trade products.

4.10. Theoretical Implications

This study made theoretical contributions to consumer research by examining the influence of information formats and characteristics on consumers' acceptance of fair- trade products. Media richness theory and the hierarchy of effects model were used to explain media effects on consumers’ cognitive, affective, and conative responses to fair- trade products. From awareness of the product to the final purchasing stage, SRCB is a complex process. This study provided insight into understanding the role that information contained in the media plays in various steps of consumer response.

4.11. Limitations and Suggestions for Future Research

146

Despite the theoretical contribution of the research, the study bears several limitations which are presented below. In addition, recommendations for future studies are offered. First, this study examined the effect of media richness by assigning participants to one of the three media formats that varied in the level of media richness and simply testing the difference in the mean scores across media formats. Similar to previous studies that treat media richness as a measurable construct (e.g., Lai & Chang,

2011; Park et al., 2008), it would be interesting to include this variable in the study’s conceptual model as a predictor of consumer responses and examine its direct and indirect influence on consumers’ acceptance of fair-trade products. By directly incorporating the media richness factor into the SEM model, the results may present a more comprehensive idea regarding the relationship between media richness and consumer responses.

Cognitive response was assessed based on individual’s objective and subjective knowledge scores. This study used a single test to measure each construct and a pre-test was conducted to simply confirm that participants’ pre-existing objective knowledge did not significantly differ across the three media format groups. However, examining the difference in the pre-test and post-test scores of both objective and subjective knowledge would provide a deeper insight into individual’s knowledge attainment after the exposure to the media content.

Another limitation of the study involves the weaknesses of using the MTurk sample. It is plausible that Mturk users are more open to and have more positive view of technology than non-users. Accordingly, their technology readiness could have affected

147 the study results. Future studies may consider using samples that better represent the population and therefore, increase the external validity of the study.

In addition, this study focused on examining the effect of information format and information characteristics which are common features of social media websites, however, represent only a limited aspect of these websites. There are many other interesting features of social media websites that convey information, such as the “Like” button, comments, or information sharing that would be meaningful to examine.

Identifying certain features that are necessary for effective communication and that significantly influence consumer behavior will present practical implications to people who are interested in social media marketing.

Furthermore, this study did not take into account diverse conditions that the viewers might face when being exposed to media postings such as the viewing time and the content of the information. In cases when the posting requires the viewer to devote more time to watch the video or read the text and when the posting content requires more cognitive involvement, the results may turn out to be different from this study. For instance, Daft and Lengel (1986) suggested that the use of rich media may not be effective in simple task situations because the rich media may create conflicting cues and distract one’s attention from the information itself. Therefore, future researchers might consider examining the influence of these additional factors which could further explain the relationship between media richness and the complexity of the content.

148

Chapter 5: Discussion

5.1. Overview

This dissertation has contributed to a better understanding of SRCB by identifying significant factors influencing consumers’ purchase decisions of socially responsible products. It consisted of three independent studies to examine various aspects of SRCB in relation to the influence of social norms and product knowledge and the ability of social media to convey purchase-relevant information. This chapter summarizes and integrates the major findings of Study 1, Study 2, and Study 3 and discusses theoretical and practical implications of the research as a whole. It concludes by addressing potential limitations of these studies and suggesting future research directions.

5.2. General Discussion

This dissertation examined diverse aspects of consumer behaviors related to consumption of socially responsible products. First, Study 1 conducted a meta-analysis of previous SRCB literature associated with a wide range of product types and ethical issues.

Next, Study 2 examined the influence of different types of social norms and product knowledge on consumers’ purchase of organic cotton and fair-trade apparel. Finally,

Study 3 examined the role of information format and information characteristics that are presented on social media websites in enhancing product knowledge and purchase intention of fair-trade products. All studies in this dissertation share a common topic of 149 identifying significant factors that affect SRCB and each study is designed to build upon and complement the findings from other studies. For example, Study 1 confirmed TPB’s strong ability in predicting SRCB by quantitatively aggregating the results from previous studies. Based on this finding, Study 2 empirically tested an extended TPB framework that incorporates different types of social norms and product knowledge. The findings from Study 2 not only provided further support for the predictive ability of TPB but also showed that the inclusion of these additional variables enhances our understanding of

SRCB. Moreover, Study 2 was designed to compare purchase behaviors between organic cotton and fair-trade apparel to complement the findings of Study 1 which had a limited number of data sets that reported results for fair-trade products. Taking into account the important role of social norms and product knowledge in the purchase process, as corroborated by the findings in the prior two studies, Study 3 examined how information presented on social media websites can increase the level of product knowledge and motivate purchase of socially responsible products. Particularly, building on the findings of Study 1 and Study 2 that offer strong theoretical support for using normative influences to encourage SRCB, Study 3 aimed at presenting practical suggestions for using social media websites, a platform where consumers interact and are exposed to others’ opinion and behaviors.

An interesting finding from Study 1 was that the subjective norms-purchase intention relationship was especially strong in the context of SRCB, a finding which was in contrast to previous TPB studies that suggested a weak association between the two

TPB constructs (i.e., subjective norms and behavioral intention). This result illustrated

150 that consumers’ intention to purchase socially responsible products is strongly linked to subjective norms; the influence of significant others in one’s social environment is important in the SRCB context. The relatively strong subjective norms-purchase intention relationship found in this study was an initial motivator for designing the latter two studies that placed emphasis on the normative influence.

Study 2 enhanced our knowledge about the effect of social norms on SRCB by distinguishing the concept into two distinct types: injunctive and descriptive norms. The findings provided an in depth understanding regarding how consumers are affected by these normative influences in regards to performing socially responsible behaviors.

Injunctive norms were likely to be more effective in encouraging positive attitudes while descriptive norms exerted a stronger influence on increasing purchase intentions (Figure

2). The two types of social norms were found to have independent effects on consumer behavior toward socially responsible apparel products.

Moreover, the results from Study 2 confirmed the crucial role of product knowledge during the consumer purchase process of socially responsible apparel products. Based on a preliminary survey, this study identified two products that consumers perceive as best representing socially responsibility in the apparel industry: organic cotton and fair-trade apparel. Although the two products were chosen by the respondents the most, the results revealed that many of the respondents felt uninformed about them; the knowledge level about organic cotton and fair-trade apparel turned out to be surprisingly low. Nevertheless, this study found that product knowledge, both objective and subjective, significantly influenced attitude toward purchasing the products.

151

Because knowledge of socially responsible apparel products is crucial in motivating

SRCB, providing consumers with more opportunities to become familiar with and objectively learn about these products would be fundamentally important. Hence, taking into account such findings, Study 3 investigated ways to increase consumers’ knowledge level and to motivate purchase related to socially responsible products, particularly by testing the effectiveness of information formats and information characteristics which are commonly features presented in the media.

First, Study 3 demonstrated a significant impact of media richness (i.e., information format) on consumers’ cognitive and affective responses. Participants in the rich media group (video format) had significantly higher scores in terms of information quality, subjective knowledge and emotion than those in the lean media group (text only format). Additionally, Study 3 found that information quantity and quality had independent effects on knowledge enhancement; information quality had an effect on both objective and subjective knowledge, whereas information quantity had an effect only on subjective knowledge. As mentioned previously, knowledge level about socially responsible products is considerably low among consumers. Thus, identifying information formats or characteristics that are more effective in enhancing consumers’ knowledge level would be especially important to people who are interested in promoting ethical consumerism.

5.3. Theoretical Implications

One of the major limitations in SRCB literature is that there is a lack of theoretical approach that explains diverse aspects of consumer behaviors related to the

152 purchase of socially responsible products. In spite of the recent increase in the awareness about ethical consumerism, extant studies provide fragmented knowledge about socially responsible consumers as well as significant factors that influence their purchase decisions.

To begin with, despite the fact that SRCB has gained attention from researchers in recent years, the use of some of the key terms in this scholarly domain seem to be inconsistent across studies in the previous literature and they often report mixed results.

For example, while many studies apply a broad definition for describing socially responsible consumers which take into account diverse ethical matters such as environmental protection, employment and human rights support, and community support, a number of studies apply a much narrower definition which involves a relatively smaller number of ethical matters. A good example of this latter case would be

Antil’s (1984) study which only considered environmental matters when profiling socially responsible consumers. Thus, clarifying the nature of socially responsible consumers by examining their purchase behaviors in a more detailed manner is necessary to provide a comprehensive understanding of this topic. In this sense, it would be useful to combine and compare results from diverse studies in the domain of SRCB. Previous literature that systematically reviews studies on socially responsible behaviors are mostly limited to pro-environmental behaviors (e.g., Aertsens et al., 2009; Bamberg & Moser,

2007) and many of them do not specifically examine consumers' purchase behaviors (e.g.,

Bamberg & Moser, 2007). To address this research gap regarding purchase behaviors of socially responsible consumers, Study 1 reviewed studies on SRCB and used meta-

153 analysis to quantitatively aggregate the results across diverse studies. It is among the first to aggregate results from SRCB studies that examine various types of product and ethical issues. Furthermore, because TPB (Ajzen, 1991) has been prevalently applied to predict

SRCB in previous studies, this theoretical framework was used as an inclusion criterion for the meta-analysis. Study 1 provided theoretical verification about the efficacy of applying TPB to explain SRCB. The statistical results of this study showed effect size values of TPB constructs that strongly support the application of TPB as an effective model for predicting SRCB.

The major contribution of both Study 2 and Study 3 was that they identified factors that facilitate purchase behaviors of socially responsible products. In particular,

Study 2 was among the first to thoroughly examine the effect of different types of social norms and product knowledge on consumers’ decision making in the context of socially responsible apparel consumption. Study 2 provides empirical evidence that social norms have two different components (i.e., injunctive and descriptive norms) and each independently affect purchase intentions for socially responsible apparel products; this finding contrasts with many previous studies that treat social norms as a single construct

(e.g., Dean et al., 2008; Kim & Han, 2010; Han et al., 2010) and suggest a weak impact of social norms on behavioral intention (e.g., Sheppard et al., 1988; Van den Putte, 1991).

These findings may help to clarify the long standing controversy about the explanatory value of social norms by highlighting the need to differentiate the concept into distinct types (Cialdini et al., 1991; Reno et al., 1993). Furthermore, both Study 2 and Study 3 demonstrated the importance of knowledge enhancement. Study 3 results revealed a

154 direct path from knowledge to consumers’ affective responses (attitude and emotion), as well as an indirect path from knowledge to purchase intention through emotion (Table

21). These findings add to our understanding of SRCB because the current literature lacks theoretical approaches that explain the relationship between knowledge and purchase behaviors of socially responsible products.

By combining the media richness theory (Daft & Lengel, 1986) and the hierarchy of effects model (Lavidge & Steiner, 1961), Study 3 empirically tested the effects of media richness on consumers’ cognitive, affective, and conative responses. Although previous literature has extensively applied these theories, only a limited number of studies have integrated the two theories to explain media effects on consumer responses

(e.g., Simon & Peppas; 2004; Xu et al., 2009). The lack of studies in this area calls for further examination considering the growing interest of academics and retailers in online marketing strategies. The results of Study 3 strongly supported the applicability of this combined model for examining the role of media information in raising knowledge about

SRCB: A significant role of media richness (i.e., information format) in consumers’ cognitive and affective responses was found and the results strongly supported the causal relationships hypothesized in the hierarchy of effects model. Furthermore, the findings of

Study 3 contribute to the theoretical development in understanding the impact of media information on consumer behavior toward socially responsible products. Study 3 confirmed that the effect of information, particularly regarding its format and characteristics (i.e., quantity and quality), needs to be considered to effectively influence consumer responses in the online environment.

155

5.4. Practical Implications

The findings of these studies present some practical implications to marketers of socially responsible products by revealing more information about their target customers.

Study 1 found that SRCB varies across product type, ethical issue, and culture. For instance, purchase intention toward socially responsible products was more strongly associated with subjective norms when buying apparel products than buying food products or hotel/tourism service products. The purchase intention-subjective norms relationship also seemed to be stronger for Asian consumers than Western consumers.

These results imply that marketers need to pay close attention to these factors as they may have significant influence on their marketing activities. Marketing strategies targeting apparel shoppers or Asian consumers may need to center upon revealing positive peer attitudes and behaviors related to the product purchase since social influence will have a relatively large impact on their purchase decision.

Study 2 further found that injunctive and descriptive norms have independent effects on SRCB. Injunctive norms had a stronger effect on forming attitudes, whereas descriptive norms more strongly affected purchase intentions. It seems ideal to highlight both types of social norms, however, in cases where it is necessary to directly increase purchase intentions of products involving diverse ethical issues, using the implications of descriptive norms would be more suitable. Companies can, therefore, highlight this aspect in their marketing strategies; perhaps, developing network-based online communities that enhance peer interactions and expose others’ socially responsible consumption habits would be an example that companies could adopt.

156

Given the importance of the social influence in motivating one’s SRCB, social media websites that provide a platform for frequent exposure to others’ opinions or behaviors would be a useful tool for marketers in this area. Accordingly, Study 3 tested the effectiveness of some of the key features of social media websites as a way to raise awareness of socially responsible consumption. Numerous companies are adopting these websites to promote sales or to enhance their brand image but currently there is not much evidence that such marketing activities are effective. Study 3 found that information format (e.g., text, text and image combined, video) had an effect on consumers’ perception about the quality of the information, subjective knowledge and emotion. The results were similar to previous studies that showed rich media have a more positive influence on consumer behaviors than lean media (e.g., Jeandrain, 2001; Park et al, 2008).

Furthermore, it demonstrated a strong influence of information quality and quantity on consumer response dimensions such as cognitive (i.e., increasing the knowledge level) and affective responses. For marketers that are adopting online promotion strategies, it would be helpful to distinguish information that has greater influence on consumer behavior, taking into account the extensive amount of information offered in the online environment these days. Empirical findings from this study regarding the effect of information formats and characteristics are expected to provide a foundation in developing effective online strategies regarding SRCB. The findings of Study 3 suggested different marketing approaches for increasing subjective and objective knowledge: centering on information quantity would be more effective for enhancement of subjective knowledge and information quality for objective knowledge. Additionally,

157 the findings indicated that inducing positive emotions toward the media content is critical because emotion has a very strong impact on purchase intention of socially responsible products. The findings suggest that video formats with detailed, accurate and timely information about ethical issues would be effective in evoking consumers’ positive emotion.

5.5. Limitations and Suggestions for Future Research

This dissertation identified important determinants of consumers’ purchase of socially responsible products which provide a sound basis for the theoretical understanding of SRCB. However, there are some limitations in this study that offers venues for future studies. To begin with, the researcher acknowledges the disadvantage of using data based on self-reports as they could be subject to reliability and validity problems. Study 2 and Study 3 used multiple items to measure each construct and assured anonymity of responses to resolve these problems; behavioral studies have reported that under these conditions self-report measures are usually both reliable and valid (Babor,

Stephens, & Marlatt, 1987). However, future research may consider using other methods that reduce social desirability bias and improve the reliability and validity of self-report data. For example, conjoint analysis involving realistic multi-attribute choice decisions is known to approximate real-world purchase decisions more accurately than simple survey items (De Pelsmacker et al., 2005).

In addition, Study 2 and Study 3 only considered Generation Y consumers who are living in the U.S. Because generation Y consumers respond positively to ethical issues (Sheahan, 2005) and are known to be consumption oriented (Sullivan & Heitmeyer

158

2008), they are likely to make up an important segment of socially responsible consumers that marketers might target. However, limiting the sample to this single generational cohort from the U.S. restricts the generalizability of the study results to a broader consumer population of other generations and from other cultures. This could be especially problematic considering the strong influence of culture on consumer behavior.

Study 1 found that SRCB significantly varies among U.S., European, and Asian consumers. For example, while subjective norms-purchase intention and PBC-purchase intention relationship were surprisingly strong for Asian consumers, these relationships were not as strong for European consumers. Thus, paying close attention to such cultural factors would be critical in generalizing the study results. Future studies could test similar research questions to a more heterogeneous consumer group that involves various types of demographics and nationalities and, perhaps, test their effect on SRCB.

Another limitation of this study is that the three studies did not examine the effect of product related attributes such as price, quality, or design of the product. These additional factors would have potential to explain SRCB or possibly moderate/mediate the relationships among the constructs. For example, because socially responsible products such as organic cotton or fair-trade products are reported to be more expensive than conventional products, testing the influence of price on purchase behaviors would have practical significance, especially to retailers who are selling these products. As seen in Appendix F, several additional items were used to measure consumers’ willingness to pay for socially responsible products in Study 2. The results showed that price is undoubtedly important to consumers when making purchase decisions. Additionally, in terms of apparel

159 shopping, aesthetic qualities such as design or style of the product would be critical attributes when consumers select the products. Therefore, future studies need to examine how product related attributes of socially responsible products influence consumer purchase intentions.

Lastly, there were some inconsistent results when examining the effect of ethical issues on consumer response regarding Study 1 and Study 2. The meta-analysis of TPB studies (Study 1) found a significant moderating role of ethical issue in explaining TPB component relationships; the strength of relationship between TPB constructs varied between products associated with environmental protections and human rights support. In contrast to this finding, TPB results regarding these two products in Study 2 turned out to be very similar. Perhaps the reason for this resulted from the difference in the product category the two studies examined. Study 1 combined results from diverse product/service categories including food, apparel, and tourism whereas Study 2 limited it to apparel products. In depth examination of whether the type of product influence consumers’ perception about ethical issues is needed to explain the contradictory result between Study 1 and Study 2.

160

References

Aertsens, J., Mondelaers, K., Verbeke, W., Buysse, J., & Van Huylenbroeck, G. (2011). The influence of subjective and objective knowledge on attitude, motivations and consumption of organic food. British Food Journal, 113(11), 1353-1378.

Aertsens, J., Verbeke, W., Mondelaars, K., & van Huylenbroeck, G. (2009). Personal determinants of organic food consumption: a review. British Food Journal, 111, 1140– 1167.

Ajzen, I. (1985). From intentions to actions: a theory of planned behavior. In J. Kuhl & J. Beckman (Eds.). Action-control: From cognition to behavior (pp. 11–39). Heidelberg: Springer.

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211.

Ajzen, I., & Fishbein, M. (1980). Understanding attitude and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall, Inc.

Ajzen, I., & Fishbein, M. (2004). Questions raised by a reasoned action approach: comment on Ogden (2003). Health Psychology, 23(4), 431-434.

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411-423.

Anuwichanont, J., Mechinda, P., Serirat, S., Lertwannawit, A., & Popaijit, N. (2011). Environmental sustainability in the Thai hotel industry. International Business & Economics Research Journal, 10(11), 91-100.

Antil, J. H. (1984). Socially responsible consumers: profile and implications for public policy. Journal of Macromarketing, 4(2), 18-39.

Armitage, C. J. & Conner, M. (2001). Efficacy of the theory of planned behaviour: A meta-analytic review. British Journal of Social Psychology (2001), 40, 471–499.

161

Arvola, A., Vassallo, M., Dean, M., Lampila, P., Saba, A., Lahteenmaki, L., & Shepherd, R.(2008). Predicting intentions to purchase organic food: The role of affective and moral attitudes in the Theory of Planned Behaviour. Appetite, 50(2-3), 443-454.

Auger, P., & Devinney, T. M. (2007). Do what consumers say matter? The misalignment of preferences with unconstrained ethical intentions. Journal of Business Ethics, 76, 361– 383.

Babor, T. F., Stephens, R. S., & Marlatt, G. A. (1987). Verbal report methods in clinical research on alcoholism: Response bias and its minimization. Journal of Studies on Alcohol, 48, 410–424.

Bailey, A. A. (2006). Retail employee theft: A theory of planned behavior perspective. International Journal of Retail & Distribution Management, 34(11), 802-816.

Bamberg, S. & Moser, G. (2007). Twenty years after Hines, Hungerford, and Tomera: A new meta-analysis of psycho-social determinants of pro-environmental behavior. Journal of Environmental Psychology, 27(1), 14-25.

Bansal, H.S. & Taylor, S.F. (2002). Investigating interactive effects in the theory of planned behavior in a service-provider switching context. Psychology & Marketing, 19(5), 407–425.

Barnes, S. J. & Vidgen, R. T. (2002). An Integrative Approach to the Assessment of E- Commerce Quality. Journal of Electronic Commerce Research, 3 (3), 114-27.

Barry, T. E. & Howard, D. J. (1990). A review and critique of the hierarchy of effects in advertising. International Journal of Advertising, 9 (2), 121-135.

Bezencon, V. & Blili, S. (2010). Ethical products and consumer involvement: What’s new? European Journal of Marketing, 44(9–10), 1305–1321.

Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: Amazon.com's Mechanical Turk. Political Analysis, 20(3), 351– 368.

Bitner, M. J. (1992). Servicescapes: the impact of physical surroundings on customers on customers and employees. Journal of Marketing, 56(2), 37-71.

Blackwell, R. D., Miniard, P. W., & Engel, J. F. (2006). Consumer Behavior 10th ed. Mason, OH: Thomson South-Western.

162

Boomsma, A. (1982). The robustness of LISREL against small sample sizesnin factor analysis models. In K.G. J¨oreskog & H. Wold (Eds.), Systems under indirect observation: Causality, structure, prediction (Part I, pp. 149–173). Amsterdam: North-Holland.

Bossinnette, M. M., & Contento, I. R. (2001). Adolescents’ perspectives and food choice behaviors in terms of the environmental impacts of food production practices: Application of a psychosocial model. Journal of Nutrition Education, 33, 72–82.

Bray, J., Johns, N., & Kilburn, D. (2011). An exploratory study into the factors impeding ethical consumption. Journal of Business Ethics, 98, 597–608.

Brown, T. J. & Dacin, P. A. (1997). The company and the product: Corporate associations and consumer product responses. Journal of Marketing, 61(January), 68-84.

Brucks, M. (1985). The effects of product class knowledge on information search behavior. Journal of Consumer Research, 12, 1-16.

Butler, S. & Francis, S. (1997). The effects of environmental attitudes on apparel purchasing behavior. Clothing and Textile Research Journal, 15, 76-85.

Carrigan, M. & De Pelsmacker, P. (2009). Will ethical consumers sustain their values in the global credit crunch?. International Marketing Review, 26(6), 674-687.

Carrigan, M., Szmigin, I., & Wright, J. (2004). Shopping for a better world? An interpretive study of the potential for ethical consumption within the older market. Journal of Consumer Marketing, 21(6), 401–417.

Carroll, A. B. (1991). The pyramid of corporate social responsibility: toward the moral management of organizational stakeholders. Business Horizons, 34, 39–48.

Chang, M. K. (1998). Predicting unethical behavior: a comparison of the theory of reasoned action and the theory of planned behavior. Journal of Business Ethics, 17(16), 1825-1834.

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative analysis. Thousand Oaks, CA: Sage.

Chen, M. F. & Tung, P. J. (2014). Developing an extended Theory of Planned Behavior model to predict consumers' intention to visit green hotels. International Journal of Hospitality Management, 36, 221–230.

Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative conduct: A theoretical refinement and reevaluation of the role of norms in human

163 behavior, In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 201-234). New York: Academic Press.

Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58, 1015-1026.

Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity and compliance. In D. T. Gilbert, S. T. Fiske, et al. (Eds.), The handbook of social psychology, Vol. 2 (4th ed., pp. 151-192). New York: McGraw-Hill.

Cochran, W. G. (1954). The combination of estimates from different experiments. Biometrics, 10, 101-129.

Cone Inc. (2013). 2013 Cone Communications social impact study. Retrieved from http://www.conecomm.com/stuff/contentmgr/files/0/e3d2eec1e15e858867a5c2b1a22c4cf b/files/2013_cone_comm_social_impact_study.pdf

Coyle, J.R. & Thorson, E.R. (2001). The effects of progressive levels of interactivity and vividness in web marketing sites. Journal of Advertising, 30(3), 65–78.

Conner, M., & Armitage, C. J. (1998). Extending the theory of planned behavior : A review and avenues for further research. Journal of Applied Social Psychology, 28, 1429- 1464.

Conner, M., Warren, R., Close, S., & Sparks, P. (1999). Alcohol consumption and the theory of planned behavior: An examination of the cognitive mediation of past behavior. Journal of Applied Social Psychology, 29, 1676–1704.

Cvijikj, I., Spiegler, E., & Michahelles, F. (2011). The effect of post type, category and posting day on user interaction level on Facebook. 2011 IEEE Third International Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third International Conference on Social Computing (pp. 810–813). Los Alamitos, CA: CPS.

Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32(5), 554–571.

Dean, M., Raats, M. M., & Shepherd, R. (2008). Moral concerns and consumer choice of fresh and processed organic foods. Journal of Applied Social Psychology, 38(8), 2088– 2107.

Demeritt, L. (2002). All Things Organic 2002: A Look at the Organic Consumer. The Hartman Group, Bellevue, WA.

164

De Pelsmacker, P., Driesen, L., & Rayp, G. (2005). Do consumers care about ethics? Willingness to pay for fair-trade coffee. The Journal of Consumer Affairs, 39(2), 363– 385.

De Pelsmacker, P. & Janssens, W. (2007). A model for fair trade buying behaviour: The role of perceived quantity and quality of information and product-specific attitudes. Journal of Business Ethics, 75, 361–380.

Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influence upon individual judgment. Journal of Abnormal and Social Psychology, 51, 629-636.

Dowd, K. & Burke, K. J. (2013). The influence of ethical values and food choice motivations on intentions to purchase sustainably sourced foods. Appetite, 1(October), 137–144.

Drolet, A., & Morrison, D. (2001). Do we really need multiple-item measures in service research? Journal of Service Research, 3(3), 196– 204.Eagry, A. H., & Chailen, S. (1993). Psychology of attitudes. Fort Worth, TX: Harcourt Brace Jovanovich.

Ellen, P. S., Mohr, L. A., &. Webb, D. J. (2000). Charitable programs and the retailer: Do they mix? Journal of Retailing, 76 (3). 393-406.

Fair Trade International (2013). What is fairtrade? Retrieved from http://www.fairtrade.net/what-is-fairtrade.html

Fair Trade USA (2013). What is fair trade? Impact. Retrieved from http://fairtradeusa.org/what-is-fair-trade/impact

Feick, L., Park, C. W., & Mothersbaugh, D. L. (1992). Knowledge and knowledge of knowledge: What we know, what we think we know and why the difference makes a difference. Advances in Consumer Research, 19, 190-192.

Ferber, R. (1977). Research by convenience. Journal of Consumer Research, 4(1), 57–58.

Ferrell, O. C., & Gresham, L. G. (1985). A contingency framework for understanding ethical decision making in marketing. Journal of Marketing, 49 (Summer), 87–96.

Fiore, A. M., Yah, X., & Yoh, E. (2000). Effects of a product display and environmental fragrancing on approach response and pleasurable experiences. Psychology and Marketing, 17(1), 27-54.

165

Fishbein, M. (1 997). Predicting, understanding, and changing socially relevant behaviors: Lessons learned. In C. McGarty & S. A. Haslam (Eds.), The message of social psychology (pp. 77-91). Oxford, UK: Blackwell.

Fitzmaurice, J. (2005). Incorporating consumers’ motivations into the theory of reasoned action. Psychology & Marketing, 22(11), 911-929.

Flynn, L. R. & Goldsmith, R. E. (1999). A short, reliable measure of subjective knowledge. Journal of Business Research, 46, 57–66.

Fornell, C., & Larker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39-50.

Frederick, E. & Webster, J. R. (1975). Determining the characteristics of the socially conscious consumer. Journal of Consumer Research, 2(3), 188-196.

Goldbach, M., Seuring, S., & Back, S. (2003). Co-ordinating sustainable cotton chains for the mass market. Greener Management International, 43, 65-79.

Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). Using social norms to motivate environmental conservation in hotels. Journal of Consumer Research, 35(3), 472-482.

Ha-Brookshire, J., & Norum, P. (2011). Cotton and sustainability: Impacting student learning through Sustainable Cotton Summit. International Journal of Sustainability in Higher Education, 12(4), 369-380.

Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate Data Analysis (7th ed.). Upper Saddle River, NJ: Pearson Prentice-Hall.

Han, H., Hsu, L., & Sheu, C. (2010). Application of the theory of planned behavior to green hotel choice: testing the effect of environmental friendly activities. Tourism Management, 31, 325–334.

Han, T. & Chung, J. (2014). Korean consumers' motivations and perceived risks toward the purchase of organic cotton apparel. Clothing and Textiles Research Journal, 32(4), 235-250.

Han, H. & Kim, Y. (2010). An investigation of green hotel customers’ decision formation: developing an extended model of the Theory of Planned Behaviour. International Journal of Hospitality Management, 29, 659-668.

166

Hauser, M., Nussbeck, F. W., & Jonas, K. (2013). The impact of food‐related values on food purchase behavior and the mediating role of attitudes: a Swiss study. Psychology & Marketing, 30(9), 765–778.

Honkanen, P. Verplanken, B., & Olsen, S.. 2006. Ethical values and motives driving organic food choice. Journal of Consumer Behavior, 5 (5), 420–430.

Hedges, L. V. & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic Press.

Hee, S. P. (2000). Relationships among attitudes and subjective norm: testing the theory of reasoned action across cultures. Communication Studies, 51(2), 162–175.

Hoch, S. J. & Ha, Y. W. (1986). Consumer learning: Advertising and the ambiguity of product experience. Journal of Consumer Research, 13 (September), 221-233.

Hofstede, G. (1980). Culture's consequences: National differences in thinking and organizing. Beverly Hills, Calif: Sage Press.

Holland, J. & Gentry, J. W. (1999). Ethnic consumer reaction to targeted marketing: a theory of intercultural accommodation. Journal of Advertising, 28(1), 65–77.

Hunter, J. E., & Schmidt, F. L. (2000). Fixed effects vs random effects meta-analysis models: Implications for cumulative research knowledge. International Journal of Selection & Assessment, 8, 275-292.

Hustvedt, G. (2006). Consumer preferences in blended organic cotton apparel. PhD dissertation, Kansas: Kansas State University.

Hustvedt, G., & Bernard, J. C. (2008). Consumer willingness to pay for sustainable apparel: The influence of labeling for fiber origin and production methods. International Journal of Consumer Studies, 32, 491–498.

Hustvedt, G., & Dickson, M.A. (2009). Consumer likelihood of purchasing organic cotton apparel: Influence of attitudes and self-identity. Journal of Fashion Marketing and Management, 13(1), 49-65.

Jaccard, J., McDonald, R., Wan, C. K., Dittus, P. J., & Quinlan, S. (2002). The accuracy of self-reports of condom use and sexual behavior. Journal of Applied Social Psychology, 32, 1863–1905.

Jeandrain, A. (2001). Consumer reactions in a realistic virtual shop: influence on buying style. Journal of Interactive Advertising, 2(10), 2-9.

167

Kaiser, F. G., Frick, J., & Stoll-Kleemann, S. (2001). On the appropriateness of self reported behavior: A validation study of the general ecological behavior scale. Diagnostica, 47, 88–95.

Kalafatis, S. P., Pollard, M., East, R., & Tsogas, M. H. (1999) Green marketing and Ajzen’s theory of planned behaviour: a cross-market examination. Journal of Consumer Marketing, 16(5), 441 - 460.

Kallgren, C. A., Reno, R. R., & Cialdini, R. B. (2000). A focus theory of normative conduct: When norms do and do not affect behavior. Personality and Social Psychology Bulletin, 26 (8), 1002-1012.

Kang, J., Liu, C., & Kim, S. (2013). Environmentally sustainable textile and apparel consumption: the role of consumer knowledge, perceived consumer effectiveness and perceived personal relevance. International Journal of Consumer Studies, 37, 442–452.

Keller, K. L. & Staelin, R. (1987). Effects of quality and quantity of information on decision effectiveness. Journal of Consumer Research, 14(2), 200-213.

Kim, H. S. & Damhorst, M. L. (1998). Environmental concern and apparel consumption. Clothing and Textiles Research Journal, 16(3), 126-133.

Kim, Y. K., Forney, J., & Arnold, E. (1997). Environmental messages in fashion advertisements. Clothing and Textiles Research Journal, 15(3), 147-154.

Kim, Y. & Han, H. (2010). Intention to pay conventional-hotel prices at a green hotel- a modification of the theory of planned behavior. Journal of Sustainable Tourism, 18(8), 997-1014.

Kim, H. & Karpova, E. (2010). Consumer attitudes toward fashion counterfeits: Application of the theory of planned behavior. Clothing and Textiles Research Journal, 28(2) 79-94.

Kim, H., Lee, E., & Hur, W. (2012). The normative social influence on eco-friendly consumer behavior: The moderating effect of environmental marketing claims. Clothing and Textiles Research Journal, 30(1), 4-18.

Kim, M. & Stoel, L. (2005). Salesperson roles: are online retailers meeting customer expectations?. Clothing and Textiles Research Journal, 33(4), 284 – 297.

Krugman, H. E. (1965). The impact of television advertising: Learning without involvement. Public Opinion Quarterly, 29, 349-356.

168

Ko, S. (2012). Predictors of purchase intention toward green apparel products in the U.S. and China. PhD dissertation, Oklahoma: Oklahoma State University.

Kozup, J., Howlett, E. H., & Pagano, M. (2008). The effects of summary information on consumer perceptions of mutual fund characteristics. Journal of Consumer Affairs, 42 (Spring), 37–59.

Lai, J-Y. & Chang, C. Y. (2011). User attitudes toward dedicated e-book readers for reading. The effects of convenience, compatibility and media richness. Online Information Review, 35 (4), 558-580.

Lavidge R. J. &Steiner C. A. (1961). A model for predictive measurements of advertising effectiveness. Journal of Marketing, 25(6), 59-62.

Lee, C. (1990). Modifying an American consumer behavior model for consumers in Confucian culture: The case of Fishbein behavioral intention model. Journal of International Consumer Marketing, 3, 27-50.

Lee, C., & Green, R. T. (1991). Cross-cultural examination of the Fishbein behavioral intention model. Journal of International Business Studies, 22, 289-305.

Li, H., Daugherty, T., & Biocca, F. (2002). Impact of 3-D advertising on product knowledge, brand attitude, and purchase intention: The mediating role of presence. Journal of Advertising, 31(3), p. 43-57.

Li, H., Daugherty, T., & Biocca, F. (2003). The role of virtual experience in consumer learning. Journal of Consumer Psychology, 13(4), 395-407.

Lin, S. (2009). Exploratory evaluation of potential and current consumers of organic cotton in Hawaii. Asia Pacific Journal of Marketing and Logistics, 21(4), 489-506.

Lin, L. Y. & Chen, C. S. (2006). The influence of the country-of-origin image, product knowledge and product involvement on consumer purchase decisions: an empirical study of insurance and catering services in Taiwan. Journal of Consumer Marketing, 23(5), 348-65.

Lipsey MW,Wilson DB. (2001). Practical meta-analysis: Applied social science research methods series. Thousand Oaks, CA: SAGE Publications.

Lodorfos, G. N. & Dennis, J. (2008). Consumers’ intent: In the organic food market. Journal of Food Products Marketing, 14, 2, 17−38.

169

Loureiro, M., L & Hine, S. (2002). A comparison of consumer willingness to pay for a local (Coloradogrown), organic, and GMO-free product. Journal of Agricultural & Applied Economics, 34(3), 477–487.

Loureiro, M. L. & Lotade, J. (2005). Do Fair Trade and eco-labels in coffee wake up the consumer conscience? Ecological Economics, 53, p. 129–138.

Ma, Y.J., Littrell, M.A., & Niehm, L. (2012). Young female consumers’ intentions toward fair trade consumption. International Journal of Retail and Distribution Management, 40 (1), 41–63.

MacGillivray, A. (2000). The fair share: The growing market share of green and ethical products. London: New Economics Foundation.

McEachern, M.G. & Warnaby, G. (2008) Exploring the relationship between consumer knowledge and purchase behaviour of value-based labels. International Journal of Consumer Studies, 32, 414–426.

Manget, J., Roche, C., & Munnich, F. (2009). Capturing the Green Advantage for Consumer Companies. Boston, MA: The Boston Consulting Group.

Mangold, W. G., & Faulds, D. J. (2009). Social media: The new hybrid element of the promotion mix. Business Horizons, 52, 357-365.

Marsh, H. W., Hau, K. T., Balla, J. R., & Grayson, D. (1998). Is more ever too much? The number of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research, 33,181–220.

McEachern, M. G., Schroder, M. J. A., Willock, J., Whitelock, J., & Mason, R. (2007). Exploring ethical brand extensions and buying behavior. Journal of Product and Brand Management, 16(3), 168-177.

McKinney, V., Kanghyun, Y., & Zahedi, F. (2002). The measurement of webconsumer satisfaction: an expectation and disconfirmation approach. Information Systems Research, 13 (3), 296–315.

Michaelidou, N. & Hassan, L. M. (2008). The role of health consciousness, food safety concern and ethical identita on attitudes and intentions towards organic food. International Journal of Consumer Studies, 32, 163–170.

Mohr, L. A. & Webb, D. J. (2005). The effects of Corporate Social responsibility and price on consumer responses. The Journal of Consumer Affairs, 39(1), 121-147.

170

Mohr, L. A., Webb, D. J., & Harris, K. E. (2001) Do consumers expect companies to be socially responsible? The impact of corporate social responsibility on buying behavior. Journal of Consumer Affairs, 35(1), 45–72.

Needham, Harper, & Steers (1985). 1985 NH&S Life Style Study, Chicago, IL.

Nilsson, J. (2008). Investment with a conscience: Examining the impact of pro-social attitudes and perceived financial performance on socially responsible investment behavior. Journal of Business Ethics, 83(2), 307-325.

Ogden, J. (2003). Some problems with social cognition models: a pragmatic & conceptual analysis. Health Psychology, 22(4), 424-428.

O’Keefe, D. J. (2002). Persuasion: Theory and research (2nd ed.). Thousand Oaks, CA: Sage. Organic Trade Associations (2013). Organic facts. Retrieved from http://www.ota.com/organic.html

Osterhus, T. L. (1997). Pro-social consumer influence strategies: When and how do they work? Journal of Marketing, 61(4), 16-29.

Park, C. W., Mothersbaugh, D. L., & Feick, L. (1994). Consumer knowledge assessment. Journal of Consumer Research, 21, 71-82.

Park, C. W. & Lessig, V.P. (1981). Familiarity and its impacts on consumer decision biases and heuristics. Journal of Consumer Research, 8, 223-230.

Park, J., Stoel, L., & Lennon, S. J. (2008). Cognitive, affective and conative responses to visual simulation: The effects of rotation in online product presentation. Journal of Consumer Behaviour, 7(1), 72–87.

Penz, E., & Stöttinger, B. (2005). Forget the "real" thing-take the copy! An explanatory model for the volitional purchase of counterfeit products. Advances in Consumer Research, 32(1), 568-575.

Peracchio, L. A. & Meyers-Levy, J. (1994). How ambiguous cropped object in ad photos can affect product evaluations. Journal of Consumer Research. 21(1), 190–204.

Pew Research Center (2010). Millennials: A portrait of generation next. Retrieved from http://www.pewsocialtrends.org/files/2010/10/millennials-confident-connected-open-to- change.pdf

171

Pieniak, Z., Aertsens, J., & Verbeke, W. (2010). Subjective and objective knowledge as determinants of organic vegetables consumption. Food Quality and Preference, 21(6), 581-588.

Rao, A and K. B. Monroe (1988). The moderating effect of prior knowledge in cue utilization in product evaluation. Journal of Consumer Research, 15, 253-263.

Ray, M. L., Sawyer, A. G., Rothschild, M. L., Heeler, R. M., Strong, E. C., & Reed, J. B. (1973). Marketing communication and the hierarchy of effects. In P. Clarke (Ed.), New models for mass communication research (pp. 147-176). Beverly Hills, CA: Sage Publishing.

Reno, R. R., Cialdini, R. B., & Kallgren, C. A. (1993). The transsituational influence of social norms. Journal of Personality and Social Psychology, 64, 104-112.

Ritchie, J., Lewis, J., & Elam, G. (2003). Designing and selecting samples. Qualitative research practice: A guide for social science students and researchers, 77-108. Rivis, A. & Sheeran, P. (2003). Descriptive norms as an additional predictor in the theory of planned behaviour: a meta-analysis. Current Psychology. 22, 218–233.

Roberts, J. A. (1995). Profiling levels of socially responsible consumer behavior: a cluster analytic approach and its implications for marketing. Journal of marketing theory and practice,3(4), 97–117.

Roberts, J. A. (1997). Sex differences in socially responsible consumers' behavior. Psychological Reports, 73, 139-148.

Robinson, R. & Smith, C. (2002). Psychology and demographic variables associated with consumer intention to purchase sustainable produced foods as defined by the Midwest Food Alliance. Journal of Nutrition Education and Behaviour, 34, 316–325.

Rosenthal, R. (1979). The “file drawer problem” and tolerance for null results. Psychological Bulletin, 86, 638–641.

Schewe, C. D., & Noble, S. M. (2000). Market segmentation by cohorts: The value and validity of cohorts in America and abroad. Journal of Marketing Management, 16(1-3), 129-142.

Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2007). The constructive, destructive, and reconstructive power of social norms. Psychological Science, 18(5), 429-434.

Schultz, W. P., Khaziana, A. M. & Zaleskia, A. C. (2008). Using normative social influence to promote conservation among hotel guests. Social Influence, 3(1), 4-23. 172

Selnes, F. & Gronhaug, K. (1986). Subjective and objective measures of product knowledge contrasted. Advances in Consumer Research, 13, 67-71.

Shaw, D., Bekin, C., Shiu, E. M. K., Hassan, L. M., Hogg, G., & Wilson, E. (2006). An examination of the volitional stages in consumer decisions to avoid sweatshop clothing. Paper presented at the meeting of the 35th European Marketing Academy Conference (EMAC), Athens, Greece.

Shaw, D., & Clarke, I. (1999). Belief formation in ethical consumer groups: An exploratory study. Marketing Intelligence & Planning, 17(2/3), 109–119.

Shaw, D. & Shui, E. (2002). An assessment of ethical obligation and self-identity in ethical consumer decision-making: A structural equation modelling approach. International Journal of Consumer Studies, 26(4), 286–293.

Shaw, D., Shiu, E., & Clarke, I. (2000). The contribution of ethical obligation and self- identity to the Theory of Planned Behaviour: An exploration of ethical consumers. Journal of Marketing Management, 16, 879–894.

Sheahan, P. (2005). Generation Y: Thriving and surviving with generation Y at work. Prahan: Hardie Grant Books.

Shepperd, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned action: A meta-analysis of past research with recommendations for modifications and future research. Journal of Consumer Research, 15, 325–343.

Simon, J. & Spero, C. P. (2004). An examination of media richness theory in product web site designs: An empirical study. Journal of Policy, Regulation and Strategy for Telecommunications, Information and Media, 6 (4), 270-281.

Sivaramakrishnan, S., Wan, F. & Tang, Z. (2007). Giving an ‘e-Human Touch’ to e- tailing: the moderating roles of static information quantity and consumption motive in the effectiveness of an anthropomorphic information agent. Journal of Interactive Marketing, 21, 1, 60–75.

Smith, R. E., Chen, J., & Yang, X. (2008). The impact of advertising creativity on the hierarchy-of-effects. Journal of Advertising, 37(4), 47–61.

Smith, J.R., Terry, D.J., Manstead, A.S.R., Louis, W.R., Kotterman, D., & Wollfs, J., (2008) The attitude–behavior relationship in consumer conduct: the role of norms, past behavior,and self-identity. Journal of Social Psychology, 148 (3), 311-334.

173

Song, H. J., Lee, C.K., Kang, S. K., & Boo, S. J. (2012). The effects of environmentally friendly perceptions on festival visitors’ decision-making process using an extended model of goal-directed behavior. Tourism Management, 22, 1417-1428.

Sparks, P., & Shepherd, R. (1992). Self-identity and the theory of planned behavior: Assessing the role of identification with “Green consumerism.” Social Psychology Quarterly, 55(4), 388-399.

Spies, K., Hesse, F., & Loesch, K. (1997). Store atmosphere, mood and purchasing behavior. International Journal of Research in Marketing, 14(1), 1-17.

Steuer, J. (1992). Defining virtual reality: Dimensions determining telepresence. Journal of Communication, 42(4), 73-93.

Sullivan, P. & Heitmeyer, J. (2008). Looking at Gen Y shopping preferences and intentions, exploring the role of experience and apparel involvement. International Journal of Consumer Studies, 32 (3), 285-295.

Sujan, M. (1985). Consumer knowledge: Effects on evaluation strategies mediating consumer judgments. Journal of Consumer Research, 12 (June), 31 -46.

Suh, K. S., & Lee, Y. E. (2005). Effects of virtual reality on consumer learning: an empirical investigation in web-based electronic commerce. MIS Quarterly, 29(4), 673- 697.

Sun, P. C. & Cheng, H. K. (2007). The design of instructional multimedia in e-Learning: a media richness theory-based approach. Computers and Education, 49(3), 662–667.

Szymanski, D. M. & Hise, R. T. (2000). E-satisfaction: An initial examination. Journal of Retailing, 76(3), 309–322.

Tarkiainen, A. & Sundqvist, S. (2005). Subjective norms, attitudes and intentions of Finnish consumers in buying organic food. British Food Journal, 107(11), 808-822.

Teng, Y. S., Wu, K. S.,& Liu, H. H. (2013). Integrating altruism and the Theory of Planned Behavior to predict patronage intention of a green hotel. Journal of Hospitality & Tourism, doi: 10.1177/1096348012471383.

Textile Exchange (2013). All about organic cotton. Retrieved from http://farmhub.textileexchange.org/learning-zone/all-about-organic-cotton.

Titus, P.A. & Bradford, J.L. (1996). Reflections on consumer sophistication and its impact on ethical business practice. Journal of Consumer Affairs, 30, 170–195.

174

Toulouse, N. O., Shiu, E., & Shaw, D. (2006). In search of fair trade: Ethical consumer decision making in France. International Journal of Consumer Studies, 30(5), 502-514.

Umberson, K. (2008). Environmentally-friendly purchase intentions: Debunking the misconception behind apathetic consumer attitudes. Texas: University of North Texas.

US Census Bureau (2013). State & county quick facts. Retrieved from http://quickfacts.census.gov/qfd/states/00000.html

US Census Bureau (2014). Latest annual retail trade report [Data file]. Retrieved from http://www.census.gov/retail/

Valacich, J., Mennecke, B., Wachter, R., & Wheeler, B. (1994). Extensions to media richness theory: A test of the task media fit hypothesis. Proceedings of the Twenty seventh Annual Hawaii International Conference on System Sciences, 4, 11-20.

Van den Putte, B. (1991). 20 years of the theory of reasoned action of Fishbein and Ajzen: A meta-analysis. Unpublished manuscript, University of Amsterdam.

Vermeir, I. & Verbeke, W. (2008). Sustainable food consumption among young adults in Belgium: Theory of Planned Behaviour and the role of confidence and values. Ecological Economics, 64, 542–553.

Vitell, S. J., Singhapakdi, A., & Thomas, J. (2001). Consumer ethics: An application and empirical testing of the Hunt-Vitell theory of ethics. Journal of Consumer Marketing, 18 (2), 153–178.

Wang, R.Y. (1998). A product perspective on total data quality management. Communications of the ACM, 41(2), 58-65.

Webster, F. E., Jr. (1975). Determining the characteristics of the socially conscious consumer. Journal of Consumer Research, 2 (December), 188-96.

Wells, W. D. (1964, October). EQ, Son of EQ, and the Reaction Profile. Journal of Marketing, 2(8), 45-52.

White, K. M., Smith, J. R., Terry, D. J., Greenslade, J. H., and McKimmie, B. M. (2009). Social influence in the theory of planned behaviour: The role of descriptive, injunctive and in-group norms. British Journal of Social Psychology, 48(1), 135–158.

Xu, H., Oh, L., & Teo, H. (2009). Perceived effectiveness of text vs. multimedia location- based advertising messaging. International Journal of Mobile Communications, 7(2),154–177.

175

Yang, X., & Smith, R. E. 2009. Beyond attention effects: Modeling the persuasive and emotional effects of advertising creativity. Marketing Science, 28(5), 935–949.

Yiridoe, E. K., Bonti-Ankomah, S., &. Martin, R. C. (2005). Comparison of consumer perceptions and preference toward organic versus conventionally produced foods: A review and update of the literature. Renewable Agriculture and Food Systems, 20(4), 193-205.

Zagata, L. (2012). Consumers’ beliefs and behavioural intentions towards organic food. Evidence from the Czech Republic. Appetite, 59, 81–89.

Zhang, Shi & Fitzsimons, G. J. (1999). Choice-process satisfaction: the influence attribute alignability and option limitation. Organizational Behavior and Human Decision Processes, 77 (3),192–214.

176

Appendix A: Study2 (Preliminary Study) Questionnaire

177

1. First Section: Introduction

Dear Participants:

I am a doctoral student of Consumer Sciences in the department of Human Sciences at the Ohio State University. The purpose of this survey is to identify the apparel products consumers typically associate with the two ethical issues (i.e., 1. environmental protection and 2. support for employment and human rights).

Please understand that your participation in this research is entirely voluntary; there will be no repercussions for non-participation. You may discontinue participation at any time and skip questions that you do not wish to answer. Even if you decide to withdraw or skip any questions, you will still receive incentives. Any information that is obtained in connection with this study will remain confidential and will be disclosed only with your permission.

If you have questions, concerns, or complaints about the study, please contact the principal investigator, Dr. Leslie Stoel, by e-mail: [email protected]. You may also contact the research staff, Tae-im Han, at [email protected]. For questions about your rights as a participant in this study or to discuss other study-related concerns or complaints with someone who is not part of the research team, you may contact Ms. Sandra Meadows in the Office of Responsible Research Practices at 1-800-678-6251. I greatly appreciate your help and you can expect to take about 2-3 minutes to participate and respond to the questionnaire. You are making a decision whether or not to participate. By clicking on the “NEXT” button below, you indicate that you have read the information provided above and you have decided to participate. If you do not wish to participate, please close this browser window.

Sincerely,

Tae-im Han, M.S. Ph.D Candidate Dept. of Human Sciences The Ohio State University Email: [email protected]

Leslie Stoel, Ph.D

178

Professor Dept. of Human Scienecs The Ohio State University Email: [email protected]

2. Second Section: Questions

A socially responsible consumer can be defined as “one who purchases products and services which he or she perceives to have a positive (or less negative) impact on the environment or uses his/her purchasing power to express current social concerns” (Roberts, 1995, p. 98).

Instructions: Please indicate your honest opinions about socially responsible consumers and products associated with ethical issues in the fashion & apparel industry.

1. Please rate your level of agreement with the following statement. Purchasing and using eco-friendly products that minimize damage to the environment would be an important concern for socially responsible consumers.  1 Strongly Disagree  2  3  4  5 Strongly Agree

2. Which product do you most relate to as supporting environmental protection in the fashion & apparel industry?  Organic cotton clothing  Recycled cotton clothing  Recycled polyester clothing  Vegetable-tanned leather clothing  Bamboo fabric clothing  Soybean fiber clothing  Others

179

3. If the product that you have in mind is not shown in the suggested answers of #2, please type your answer here.

______

4. Please rate your level of agreement with the following statement. Purchasing and using products that support fair labor practices and human rights would be an important concern for socially responsible consumers.  1 Strongly Disagree  2  3  4  5 Strongly Agree

5. Which product do you most relate to as supporting the improvement of employment and human rights conditions in the fashion & apparel industry?  Sweatshop free clothing  Fair trade clothing  Child labor free clothing  Union made clothing  Others

6. If the product that you have in mind for #5 is not shown in the suggested answers, please type your answer here.

______

180

Appendix B: Study2 (Main Study) Questionnaire

181

Dear Participants:

I am a doctoral student of Consumer Sciences in the department of Human Sciences at th e Ohio State University. The purpose of this research is to contribute to a better understanding of socially responsible consumers by identifying significant factors influencing their purchase decisions.

If you have questions, concerns, or complaints about the study, please contact the principal investigator, Dr. Leslie Stoel, by e-mail: [email protected]. You may also contact the research staff, Tae-im Han, at [email protected]. For questions about your rights as a participant in this study or to discuss other study-related concerns or complaints with someone who is not part of the research team, you may contact Ms. Sandra Meadows in the Office of Responsible Research Practices at 1-800-678-6251. I greatly appreciate your help and you can expect to take about 10 minutes to participate and respond to the questionnaire. You are making a decision whether or not to participate. By clicking on the “NEXT” button below, you indicate that you have read the information provided above and you have decided to participate. If you do not wish to participate, please close this browser window.

Sincerely, Tae-im Han, M.S. Ph.D Candidate Dept. of Human Sciences The Ohio State University Email: [email protected]

Leslie Stoel, Ph.D Professor Dept. of Human Scienecs The Ohio State University Email: [email protected] 182

2. Second Section: Questions

Is your age between 20-37?  Yes  No

Please indicate the level of agreement you have with each statement. 1- 2 3 4- 5 6 7- Strongly Neither Strongly Disagree Agree Agree nor Disagree I know a lot about organic cotton        apparel.

I feel knowledgeable about organic cotton        apparel.

Among my circle of friends, I’m one of the “experts” on organic        cotton apparel.

Compared to most other people, I know        more about organic cotton apparel.

183

Please choose whether you think each of the following statements is true or false. After each answer, indicate your degree of certainty that your answer will be marked as correct. I think that the How certain are you that this answer is statement is... right? 1-Very 5- Very True False 2 3 4 Uncertain Certain Organic cotton is produced without the use        of synthetic pesticides.

Organic cotton farmers may use synthetic        fertilizers.

Organic cotton farmers may use genetically        modified seeds.

The Global Organic Textile Standard (GOTS) certification indicates that chemical inputs during post-harvest processing (e.g., dyes and process chemicals) have been assessed and meet        certain toxicological and environmental criteria.

How would you describe your attitude toward buying organic cotton apparel? For me buying organic cotton apparel would be: 1 2 3 4 5 6 7 Negative        Positive Unpleasant        Pleasant Foolish        Wise A bad idea        A good idea Undesirable        Desirable

184

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree People who influence my decisions would approve of me buying organic        cotton apparel.

People who are important in my life would approve of me buying organic cotton        apparel.

Close friends and family think it is a good idea for        me to purchase organic cotton apparel.

Please indicate your opinions about the following questions.

How many of the people who are important to you would buy organic cotton apparel in the near future?  1- None  2  3  4- Quite a Bit  5  6  7- All

What proportion of the people who are important to you buy organic cotton apparel?  1- None  2  3  4- Quite a Bit  5  6  7- All

185

How likely is it that people who are important to you buy organic cotton apparel?  1- Extremely Unlikely  2  3  4- Neutral  5  6  7- Extremely Likely

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I believe that I have the resources and the ability to purchase organic        cotton apparel.

I do not face high barriers in purchasing        organic cotton apparel.

If I wanted to, I could easily buy organic        cotton apparel.

How much control do you think you have over purchasing organic cotton apparel in the near future?  1- Absolutely No Control  2  3  4- Neutral  5  6  7- Complete Control

186

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I would like to purchase organic cotton apparel        in the future.

If I see organic cotton apparel, I intend to purchase or consider        purchasing it.

If I see a retail store selling organic cotton apparel, I intend to visit        the store to purchase a product.

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I would go out of my way to find organic cotton        apparel to purchase.

I would purchase organic cotton apparel only if the price is the        same or lower than other apparel.

How much more will you pay for organic cotton apparel compared with conventional products?  I will not pay more for organic cotton apparel.  less than 5%  5-10%  10-20%  20-40%  More than 40%

187

How many times have you purchased organic cotton clothing?  0 time  1 time  2~3 times  4~5 times  more than 6 times

Have you purchased organic products other than organic cotton clothing?  Yes  No

If yes, what kind of product have you purchased?  Organic Food  Other Organic Fabric Clothing  Organic Cosmetic and Skin Care Product  Others ______

Please indicate the level of agreement you have with each statement 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I know a lot about fair- trade apparel.       

I feel knowledgeable about fair-trade apparel.       

Among my circle of friends, I’m one of the “experts” on fair-trade        apparel.

Compared to most other people, I know more        about fair-trade apparel.

188

Please choose whether you think each of the following statements is true or false. After each answer, indicate your degree of certainty that your answer will be marked as correct. I think that the How certain are you that this answer is right? statement is... 1-Very 5-Very True False 2 3 4 Uncertain Certain Fair-trade certified labels indicate that the products are also        certified organic.

Fair-trade avoids direct trade between importers and fair-        trade producers.

Fair-trade prohibits the use of GMOs (Genetically Modified        Organisms).

Fair-trade prohibits forced child and slave        labor.

How would you describe your attitude toward buying fair-trade apparel? For me buying fair-trade apparel would be: 1 2 3 4 5 6 7 Negative        Positive Unpleasant        Pleasant Foolish        Wise A bad idea        A good idea Undesirable        Desirable

189

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree People who influence my decisions would approve of me buying fair-trade        apparel.

People who are important in my life would approve of me buying fair-trade        apparel.

Close friends and family think it is a good idea for        me to purchase fair-trade apparel.

Please indicate your opinions about the following questions.

How many of the people who are important to you would buy fair-trade apparel in the near future?  1- None  2  3  4- Quite a Bit  5  6  7- All

What proportion of the people who are important to you buy fair-trade apparel?  1- None  2  3  4- Quite a Bit  5  6  7- All

190

How likely is it that people who are important to you buy fair-trade apparel?  1- Extremely Unlikely  2  3  4- Neutral  5  6  7- Extremely Likely

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I believe that I have the resources and the ability to purchase fair-trade        apparel.

I do not face high barriers in purchasing        fair-trade apparel.

If I wanted to, I could easily buy fair-trade        apparel.

How much control do you think you have over purchasing fair-trade apparel in the near future?  1- Absolutely No Control  2  3  4- Neutral  5  6  7- Complete Control

191

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I would like to purchase fair-trade apparel in the        future.

If I see fair-trade apparel, I intend to purchase or consider        purchasing it.

If I see a retail store selling fair-trade apparel, I intend to visit        the store to purchase a product.

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I would go out of my way to find fair-trade apparel        to purchase.

I would purchase fair- trade apparel only if the        price is the same or lower than other apparel.

How much more will you pay for fair-trade apparel compared with conventional products?  I will not pay more for fair-trade apparel.  less than 5%  5-10%  10-20%  20-40%  More than 40%

192

How many times have you purchased fair-trade clothing?  0 time  1 time  2~3 times  4~5 times  more than 6 times

Have you purchased fair-trade products other than fair- trade clothing?  Yes  No

If yes, what kind of product have you purchased?  Fair-trade Food  Fair-trade Cosmetic and Skin Care Product  Others ______

In what year were you born? ______

Your race/ethnicity is:  White  Black or African American  Hispanic or Latino  American Indian or Alaska Native  Asian  Other ______

Your gender is:  Male  Female

What is the highest level of education you have attained?  Did Not Complete High School  High School/GED  Some College  Associate Degree  Bachelor's Degree  Master's Degree  Advanced Graduate/Professional work or Ph.D.  Other ______

193

What is your annual personal income?  Less than $10,000  $10,000 - $24,999  $25,000 - $34,999  $35,000 - $49,999  $50,000 to $74,999  More than $75,000

194

Appendix C: Study3 Questionnaire

195

1. First Section: Introduction

Dear Participants:

I am a doctoral student of Consumer Sciences in the department of Human Sciences at the Ohio State University. The purpose of this research is to examine the effect of product knowledge on purchase intentions and test the effectiveness of using social media websites to educate and promote ethical consumer behaviors.

Please understand that your participation in this research is entirely voluntary; there will be no repercussions for non-participation. You may discontinue participation at any time and skip questions that you do not wish to answer. Even if you decide to withdraw or skip any questions, you will still receive incentives from the survey firm. Any information that is obtained in connection with this study will remain confidential and will be disclosed only with your permission.

If you have questions, concerns, or complaints about the study, please contact the principal investigator, Dr. Leslie Stoel, by e-mail: [email protected]. You may also contact the research staff, Tae-im Han, at [email protected]. For questions about your rights as a participant in this study or to discuss other study-related concerns or complaints with someone who is not part of the research team, you may contact Ms. Sandra Meadows in the Office of Responsible Research Practices at 1-800-678-6251. I greatly appreciate your help and you can expect to take about 10-15 minutes to participate and respond to the questionnaire. You are making a decision whether or not to participate. By clicking on the “NEXT” button below, you indicate that you have read the information provided above and you have decided to participate. If you do not wish to participate, please close this browser window.

Sincerely,

Tae-im Han, M.S. Ph.D Candidate Dept. of Human Sciences The Ohio State University Email: [email protected]

Leslie Stoel, Ph.D Professor

196

Dept. of Human Scienecs The Ohio State University Email: [email protected]

2. Second Section: Questions

Please choose whether you think each of the following statements is true or false. After each answer, indicate your degree of certainty that your answer will be marked as correct. I think that the How certain are you that this answer is statement is... right? 1-Very 5-Very  True False 2 3 4 Uncertain Certain Fair-trade certified labels indicate that the products are also        certified organic.

Fair-trade avoids direct trade between importers and fair-        trade producers.

Fair-trade prohibits the use of GMOs (Genetically Modified        Organisms).

Fair-trade prohibits forced child and slave        labor.

197

How would you describe your attitude toward buying fair- trade products? For me buying fair-trade products would be: 1 2 3 4 5 6 7 Negative        Positive Unpleasant        Pleasant Foolish        Wise A bad idea        A good idea Undesirable        Desirable

Imagine that you are viewing a posting containing the following article on a social media website. Please read the article: After reading the article, press the next button below. You need to spend at least 2 minutes reading the article to move on to the next page. If you do not press the next button in 10 minutes, it will advance automatically to the next page of the survey.

Click here to go to the article.

198

The following statements are about the posting that you just saw. Please choose the answer that you believe is true for each statement. Yes No Not Sure The posting    included images. The posting included moving    images. The posting    included sounds.

Please indicate the level of agreement you have with each statement. Information contained on the posting … 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree provides accurate information.       

provides believable information.       

provides current and timely information.       

provides relevant information.       

presents the information in an        appropriate format.

199

In my opinion, the information about fair-trade provided by the posting was...  1- Very Brief  2  3  4- Neutral  5  6  7- Very Detailed

The amount of information about fair-trade in the posting was...  1- A Little  2  3  4- Neutral  5  6  7- A Lot

The information about fair-trade provided by the posting seems...  1- Not At All Complete  2  3  4- Neutral  5  6  Very Complete

200

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I know a lot about fair- trade.       

I feel knowledgeable about fair-trade.       

Among my circle of friends, I’m one of the        “experts” on fair-trade.

Compared to most other people, I know        more about fair-trade.

201

Please choose whether you think each of the following statements is true or false. After each answer, indicate your degree of certainty that your answer will be marked as correct. I think that the How certain are you that this answer is statement is... right? 1-Very 5-Very  True False 2 3 4 Uncertain Certain Fair-trade offers improved terms of trade and better prices to workers and        farmers in developing countries

Fair-trade helps farmers and workers in developing countries to obtain business skills needed to produce        high-quality products so that they can compete in the global marketplace.

Fair-trade encourages importers to purchase fair-trade products        through middlemen.

Fair-trade principles do not include        environmental sustainability.

202

Please indicate the level of agreement you have with each statement regarding the posting that you just saw. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree This posting is very        appealing to me. This is a heart-        warming posting. I like this posting.        This posting makes        me feel good. This is a wonderful        posting. This is a fascinating        posting.

Please indicate the level of agreement you have with each statement. 1- 2 3 4 5 6 7- Strongly Strongly Disagree Agree I would like to purchase fair-trade        products in the future.

If I see fair-trade products, I intend to purchase or consider        purchasing it.

If I see a retail store selling fair-trade products, I intend to        visit the store to purchase a product.

In what year were you born? ______

203

Your race/ethnicity is:  White  Black or African American  Hispanic or Latino  American Indian or Alaska Native  Asian  Other ______

Your gender is:  Male  Female

What is your annual personal income?  Less than $10,000  $10,000 - $24,999  $25,000 - $34,999  $35,000 - $49,999  $50,000 to $74,999  More than $75,000

Have you ever purchased a fair-trade product?  Yes  No

How many times have you purchased fair-trade products within the last 12 months?  0 time  1 time  2~3 times  4~5 times  more than 6 times

204

If you have previously purchased fair-trade products, what kind of product was it?  Fair-trade Food  Fair-trade Clothing  Fair-trade Cosmetic and Skin Care Product  Others ______

205

Appendix D: Text +Image Posting

206

207 and be reliable business partners to US companies that want to buy their products. So, we see a lot of the fair-trade premium money being spent by co-ops strengthening their own quality and business capacity. I was in Rwanda recently visiting coffee farms that are part of the fair-trade community; they built new mills, trained their farmers how to produce in better quality, and built cupping laboratories that allow the farmers to learn the taste of their own products for the first time. I was very inspired to see how the Rwandan farmers have come together into co-ops and built the foundation for producing some of the most delicious coffee in the world. What we have discovered is that there is a direct correlation between the quality of a food product and the amount of money the farmers receive. So, if we ensure that farmers get a decent return to their harvest, they are able to produce in much higher quality. It is just logical and the difference is there in all the contests, awards and competitions that the fair-trade farmers continuously win every year.

208

Appendix E: Text Only Posting

209

Fair-trade is all about improving lives, but we don't do that through charity. There is no hand out in the Fair Trade movement. In so many countries in Africa, Asia and Latin America, the laws protecting workers are very weak or they're not enforced. For much of the agricultural working population around the world, life is very hard and not very safe. That is what fair trade is trying to change. In Ghana, for example, historically, child labor has been a huge problem. We found that fair-trade in Cocoa co-ops are building schools, creating scholarship funds, and sending kids to high schools and to colleges. In Rwanda, we found fair-trade communities that have invested in clean water and women's health. In Uganda, vanilla farmers get 8% of the final export price on average, but by connecting with the US fair-trade buyers, they have been able to get a much higher price because they're not selling to local middle men. With the higher price, they in turn have been able to invest in the quality of the product and the vanilla now is just incredible. The biggest change that we see in large fair-trade farms and apparel factories is that people are working in safer conditions, getting bonuses, and are able to address problems with the management. So there is a level of confidence that you don't see outside in other workplaces. There is an invisible dividend in the fair-trade world that you cannot see but know when you talk to a farmer. It’s hope, pride and dignity because people are solving their own problems through fair trade. Fair-trade is not a charity. American consumers are not going to buy fair-trade products just because it helps some farmers. If farmers get fair-trade certified but don't produce a good product, it's not going to get sold in the market. One of the important features a fair trade is that we are helping farmers invest in quality because their long- term income is very closely connected with their ability to produce high-quality products and be reliable business partners to US companies that want to buy their product. So, we see a lot of the fair-trade premium money being spent by co-ops strengthening their own quality and business capacity. I was in Rwanda recently visiting coffee farms that are part of the fair-trade community; they built new mills, trained their farmers how to produce in better quality, and built cupping laboratories that allow the farmers to learn the taste of their own products for the first time. I was very inspired to see how the Rwandan farmers have come together into co-ops and built the foundation for producing some of the most delicious coffee in the world. What we have discovered is that there is a direct correlation between the quality of a food product and the amount of money the farmers receive. So, if we ensure that farmers get a decent return to their harvest, they are able to produce in much higher quality. It is just logical and the difference is there in all the contests, awards and competitions that the fair-trade farmers continuously win every year.

210

Appendix F: Willingness to Pay and Willingness to Search

211

Organic cotton Fair-trade Total I will not pay more for 76 70 146 (29.2%) organic cotton apparel. less than 5% 79 60 139 (27.8%)

5-10% 69 84 153 (30.6%) 10-20% 24 30 54 (10.8%) 20-40% 2 6 8 (1.6%) Responses to the item “How much more will you pay for organic cotton (fair-trade) apparel compared with conventional products?”

Organic cotton Fair-trade Total 1-Strongly Disagree 9 10 19 (3.8%) 2 11 10 21 (4.2%) 3 21 32 53 (10.6) 4- Neither Agree nor 43 48 91 (18.2%) Disagree 5 51 64 115 (23%) 6 68 44 112 (22.4%) 7- Strongly Agree 47 42 89 (17.8%) Responses to the item “I would purchase organic cotton (fair-trade) apparel only if the price is the same or lower than other apparel”

Organic cotton Fair-trade Total 1-Strongly Disagree 59 41 100 (20%) 2 50 45 95 (19%) 3 52 43 95 (19%) 4- Neither Agree nor 38 63 101 (20.2%) Disagree 5 35 35 70 (14%) 6 10 17 27 (5.4%) 7- Strongly Agree 6 6 12 (2.4%) Responses to the item “I would go out of my way to find organic cotton (fair-trade) apparel to purchase”.

While examining these items were not part of the main analysis, they were

212 included in the questionnaire to get an overall picture of consumers’ willingness to pay a premium price and willingness to go out of their way to search because organic cotton and fair-trade apparel products are generally more expensive and less available than conventional products. The results showed that 43% of the consumers are willing to pay at least 5% more for these products. However, 63.2% of the same respondents indicated that they would purchase the products only if the price is equal to or lower than other conventional products. Also, only 21.8% agreed that they would go out of their way to find organic cotton or fair-trade apparel to purchase. The willingness to pay more scale demonstrates that that demand for apparel products with ethical labels does exist in the current market. Yet, further results provide evidence that high price and limited availability would be important factors causing them to feel reluctant to purchase these products.

As seen in the following figure, the TPB models of organic cotton and fair-trade apparel are tested after consumers’ willingness to pay more for the products has been controlled. In both cases, the willingness to pay more and purchase intention had strong association; the more likely to be willing to pay more for the products, the stronger the purchase intention. The results of these models were similar to that of the original models from Study 2. Attitude, PBC, and descriptive norms were significant determinants of intention to purchase organic cotton apparel and attitude, injunctive norms, and descriptive norms were significant determinants of intention to purchase fair-trade apparel, after controlling for the effect of the willingness to pay more. One slight difference when adding the control variable is that the effect of PBC on purchase

213 intention became weaker; this effect became insignificant in the fair-trade model.

Although these additional items were examined here, they did not provide a detailed explanation on how the relatively high price of the products affects consumers’ purchase decisions. Because apparel products with organic or fair-trade labels are usually sold in a higher price range, it would be important to closely study the influence of price.

Future studies may further investigate how the construct, willingness to pay more, is actually associated with purchase intentions and also test its relationship with other constructs included in this study.

Note. All values are standardized estimates. Significant effect No effect, ** p<.01, * p<.05.

Summary of SEM Results Controlling for the Effects of Willingness to Pay More

214