ENACTED AFFORDANCES OF SOCIAL MEDIA AND CONSUMERS’ RESPONSE TO ADVERTISING By Jing Yang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Information and Media—Doctor of Philosophy 2017 ABSTRACT ENACTED AFFORDANCES OF SOCIAL MEDIA AND CONSUMERS’ RESPONSE TO ADVERTISING By Jing Yang With the rise of diversification of individuals’ social media use, the need for understanding the differences across various social media platforms has become more salient. However, based on the findings of a 10-year-long literature review of social media advertising, prior scholars rarely focused on the medium differences in social media, as compared to other factors, such as message variance, individual differences, and source influence. To fill such a research gap, the fundamental purpose of this research project, thus, is to introduce the new concept titled enacted affordances of social media and its typologies to understand the differences across various social media platforms. Additionally, the current research project also holds the purpose to showcase how the newly developed concept can be applied in empirical communication studies. Therefore, taking social media advertising context as an example, this project investigated the relationship between the enacted affordances of social media and individuals’ responses to social media advertising. Specifically, the study compared the six types of enacted affordances of social media regarding their influential impact on advertising acceptance and avoidance. Results indicated that, in general, the enacted affordance of non-relational content had a stronger impact than the enacted affordance of relational content in driving consumers’ acceptance of social media advertising. The enacted affordance of content creation of a social media platform had more impact than the enacted affordance of content consumption in increasing consumers’ acceptance of social media advertising. Moreover, the current study also suggests that consumers’ advertising avoidance and advertising acceptance are not two sides of one coin due to the fact that the enacted affordances of social media were not found having the same significant relationships with advertising avoidance as advertising acceptance. The research project ends with discussion its theoretical contribution, managerial implications, as well as the limitations and future research directions. Copyright by JING YANG 2017 This dissertation is dedicated to my soulmates, Zhaoqin Zhang and Zheng Yuan, for standing by my side and providing endless support during the completion of this dissertation. Also, this dissertation is dedicated to God. I could never be strong and courageous enough to finish this process without the love and power from you. Thank you, my lord, for being the light in the darkness and giving me faith in you. ACKNOWLEDGEMENTS This dissertation could not be completed without the help of many people. They have been role models in my life, showed me the integrity and passion for research, for teaching, for life, and for love. First of all, I would like to express my sincere gratitude to my advisor, Dr. Hairong Li, who has been my mentor ever since my master program back in 2010 at Michigan State University. Over the past seven years, his scholarship and vision have been so inspiring to me that lead me to my passion for digital advertising. I could never have grown this much as a scholar and an educator without the support, care, and examples from Dr. Li. I am sincerely grateful for the amount of time and guidance offered by Dr. Li all along the way of my Ph.D. program. Meanwhile, I also want to thank my awesome committee members, Dr. Rabindra Ratan, Dr. Constantinos K. Coursaris, Dr. Wietske Van Osch, the top-edge scholars I met here at Michigan State University. They have shown me the integrity, the sincerity of being a scholar and provided my tremendous intellectual and emotional support in the process of the completion of my dissertation. I thank each one of you, and can’t be more grateful for what I’ve received from you. Moreover, special thanks to my dearest friends, Yingqian Lin, Xiaoyu Zhao, Tong Wu, and all my brothers and sisters at Lansing Chinese Christian Church. For all the ups and downs of my Ph.D. life, you’ve been there for me. Thank you for your empathy when I was in sorrow, thank you for your encouragement when I was weary, and thank you for your prayers when I desperately needed it. Thank you, for being my constant companion, all these years. i TABLE OF CONTENT LIST OF TABLES ......................................................................................................................... iv LIST OF FIGURES ....................................................................................................................... iv CHAPTER 1 ................................................................................................................................... 1 INTRODUCTION.......................................................................................................................... 1 CHAPTER 2 ................................................................................................................................... 6 REVIEW OF SOCIAL MEDIA ADVERTISING......................................................................... 6 Medium-factor Studies in Social Media Advertising ................................................................. 8 Comparison of Social Media with Traditional Media ............................................................ 9 Comparison of Different Social Media ................................................................................. 10 Individuals’ Relationship with Medium ............................................................................... 10 Macro Analyses of Social Medium....................................................................................... 11 Conclusion ................................................................................................................................ 12 CHAPTER 3 ................................................................................................................................. 14 A HYBRID APPROACH OF MEDIUM AND INDIVIDUAL: ENACTED AFFORDANCES OF SOCIAL MEDIA.................................................................................................................... 14 Affordances, Perceived Affordances and Enacted Affordances ............................................... 15 Enacted Affordances of Social Media ...................................................................................... 18 Media Characteristics and Social Media Use ....................................................................... 18 Individual Characteristics and Social Media Use ................................................................. 19 Typology of Enacted Affordances of Social Media ............................................................. 20 Social vs. Non-Social Goals of Social Media Use ................................................................ 21 “Consuming”, “Contributing” and “Creating” on Social Media .......................................... 23 CHAPTER 4 ................................................................................................................................. 28 ENACTED AFFORDANCES AND CONSUMERS' RESPONSE TO SOCIAL MEDIA ADVERTISING.............................................................................................................................28 Social Media Advertising: A Paid Format of Marketing Communication ............................... 28 Consumer Acceptance and Avoidance of Social Media Advertising ....................................... 30 Consumer Acceptance of Social Media Advertising ............................................................ 30 Consumer Avoidance of Social Media Advertising ............................................................. 32 Enacted Affordances of Social Media and Consumer Responses to Advertising .................... 34 Enacted Affordances of Social Media as Acquired Knowledge ........................................... 35 Contextual Congruence and Consumer Response ................................................................ 38 Cognitive Capacity in Dual-Process Information Processing and Consumer Response ...... 41 CHAPTER 5 ................................................................................................................................. 45 ii METHOD......................................................................................................................................45 Study Design ............................................................................................................................. 45 Measurement ............................................................................................................................. 46 Sample Description ................................................................................................................... 48 Sample’s Social Media Selection Description .......................................................................... 49 CHAPTER 6 ................................................................................................................................. 50 DATA ANALYSIS AND HYPOTHESES TESTING................................................................. 50 Factor Validity and Reliability Test .......................................................................................... 50 Enacted Affordances Differences of Social Media Platforms .................................................. 51 Advertising Acceptance and Avoidance Differences Across Social Media Platforms ............ 52 Statistical Analysis Assumption Examination .......................................................................... 53 Hypotheses Testing ................................................................................................................... 55 Analysis Approach 1: Enacted Affordances as Individual Measures ................................... 56 Analysis Approach 2: Enacted Affordances as Calculated Area .......................................... 63 CHAPTER 7 ................................................................................................................................. 70 DISCUSSION & CONCLUSION............................................................................................... 70 CHAPTER 8 ................................................................................................................................. 76 CONTRIBUTIONS, LIMITATIONS, AND FUTURE DIRECTION..........................................76 APPENDIX ................................................................................................................................... 80 REFERENCES ........................................................................................................................... 118 iii LIST OF TABLES Table 1. Statistics of Published Social Media Advertising Research from 2006 – 2016………. 81 Table 2. Thematic Categorization of Social Media Advertising Studies from 2006 - 2016….... 82 Table 3. Typologies of Enacted Affordances of Social Media ……………………………….... 84 Table 4. Finalized Scale Items for Enacted Affordances of Social Media………………………85 Table 5. Measurement Table ……………………………….…………………………………... 86 Table 6. Sample Description……………………………….………………………………….....87 Table 7. Selection/Use of Social Media Platforms……….………………………………….......88 Table 8. Factor Analysis and Reliability Test …………….…………………………………......89 Table 9. Enacted Affordances of Social Media Platforms.…………………………………........90 Table 10. ANOVA Analysis Result for Advertising Acceptance and Avoidance………………93 Table 11. Correlation Analysis for Independent Variables.……………………………………. 96 Table 12. Standardized Coefficients and Confidence Interval Results_H1……………………. 97 Table 13. Standardized Coefficients and Confidence Interval Results_H2…………………….101 Table 14. Standardized Coefficients and Confidence Interval Results_H3……………………102 Table 15. Standardized Coefficients and Confidence Interval Results_H4……………………106 Table 16. Standardized Coefficients and Confidence Interval Results_Area_H_Accept……...108 Table 17. Standardized Coefficients and Confidence Interval Results_Area_H_Avoid……….109 Table 18. Standardized Coefficients and Confidence Interval Results_Area_All_Accept…….113 Table 19. Standardized Coefficients and Confidence Interval Results_Area_All_Avoid……...117 iv LIST OF FIGURES Figure 1. Enacted Affordances for Each Social Media Platform……………………………….. 92 Figure 2. Advertising Acceptance Across Social Media Platforms…………………………….. 94 Figure 3. Beta Coefficient Comparison using 95% Confidential Intervals…………………….. 95 Figure 4a. Beta Coefficient Comparison of Enacted Affordances of Relational Content Consumption and Non-relational Content consumption using 95% Confidential Intervals…… 98 Figure 4b. Beta Coefficient Comparison of Enacted Affordances of Relational Content Contribution and Non-relational Content Contribution using 95% Confidential Intervals.…… 99 Figure 4c. Beta Coefficient Comparison of Enacted Affordances of Relational Content Creation and Non-relational Content Creation using 95% Confidential Intervals.…….…….…………..100 Figure 5a. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Consumption and Contribution using 95% Confidential Intervals.…….…….…………...……103 Figure 5b. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Consumption and Creation using 95% Confidential Intervals.…….…….…………...…….….104 Figure 5c. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Contribution and Creation using 95% Confidential Intervals.…….…….…………...…….…...105 Figure 6. Example for the Computation of Enacted Affordance Areas.…….…….…………....106 Figure 7a. Beta Coefficient Comparison of Enacted Affordances of Relational Content and Nonrelational Content using 95% Confidential Intervals.…….…….…………....…….…….……..110 Figure 7b. Beta Coefficient Comparison of Enacted Affordances of Content Consumption and Content Contribution under Heuristic Condition using 95% Confidential Intervals…….…......111 Figure 7c. Beta Coefficient Comparison of Enacted Affordances of Content Contribution and Content Creation under Heuristic Condition using 95% Confidential Intervals…….…….…...112 Figure 8a. Beta Coefficient Comparison of Enacted Affordances of Content Consumption and Content Contribution using 95% Confidential Intervals.…….…….…………...…….….…….114 Figure 8b. Beta Coefficient Comparison of Enacted Affordances of Content Creation and Content Consumption using 95% Confidential Intervals.…….…….…………...…….….……115 Figure 8c. Beta Coefficient Comparison of Enacted Affordances of Content Contribution and Content Creation using 95% Confidential Intervals.…….…….…………...…….….….....116 v CHAPTER 1 INTRODUCTION Social media has become one of the major media repertoire used in people’s daily life, as indicated in Pew Research report that more than two-thirds of United States adults are using at least one type of social media as of Nov. 2016 (Pew Research, 2017). Defined as “a group of Internet-based applications that build on ideological and technological foundations of Web 2.0, which allows the creation and exchange of user-generated content” (Kaplan and Haenlein, 2010), social media has now been presented in various forms of interaction, communication, collaboration, and community formation (Zhao et al., 2013). And the evolvement of various features presented in different social media platforms is providing users different kinds of digital experiences on these social media platforms. The growing diversification of social media not only brought excitement to the users but also to the media industry and the academia for the opportunities to develop novel social media campaigns and new knowledge about consumer responses to various social media advertising. Prior literature of social media advertising has witnessed the evolvement of social media advertising, from early forms of electronic word-of-mouth (e.g., Smith et al. 2007; Dwyer, 2007; Prendergast et al. 2010), brand community (e.g., Casaló et al. 2008; Schau et al. 2009), viral advertising (e.g., Hinz et al. 2011; Van Noort et al. 2012) and consumer engagement (e.g., Muntinga et al. 2011; Tsai & Men 2013; Holleeek et al. 2013), to the latest targeted sponsored content and consumer conversion (e.g., Keyzer et al. 2015; Van Reijmersdal et al. 2016). The development of social media advertising research is quite synchronized as the development of the industry’s advertising products, which has gradually been divided into owned, earned and paid 1 advertising content. Moreover, scholars have also investigated various aspects of these different topics in social media advertising. Specifically, based on the classic communication model (Lasswell 1948; Stern 1994), the researcher categorized the variables examined in empirical studies into studies focusing on source factors (message sender), individual factors (message receiver), message factors and medium factors. After reviewing 133 empirical studies published in top journals in advertising and marketing field during 2006 - 2016, the results revealed that majority (67%) of the studies focused on individual factors such as one’s motivation in engaging with brand in social media (Muk & Chung 2014; Kwon et al. 2014), or one’s individual characteristics such as gender difference (Hoy & Milne 2010; Do et al. 2014; Kim and Yoon 2014), just to name a few. A fair amount of studies has also investigated the source effects (26%) such as perceived social strength (e.g., Pan and Chiou 2011; Chu and Kim 2011) and message factors (34%) such as message valence (e.g., Corstjens & Umblij 2012; Daughtery & Hoffman 2014). However, only a few studies (10%) were conducted to understand the medium effect in social media advertising. Thus, the researcher suggests that more effort should be devoted to the medium-approach in understanding social media advertising effectiveness. The medium approach in understanding communication process was coined by Marshall McLuhan’s axiom “The Medium is the Message,” in which he suggested that the medium itself “shapes and controls the scale and form of human association and action” (Understanding Media. NY, 1964, p.9). The underlying derivation of such axiom considered the medium characteristics are having unique impact on individuals’ attitude, cognitive and behavioral reactions toward the messages embedded in the medium. In other words, independent from the messages factors, the 2 medium itself generates effects upon message recipients’ responses (McLuhan & Fiore, 1967). Although acknowledging the importance of medium factors in communication studies, theories and studies in medium effects have always been under-development as compared to other major communication factors, such as source factors, individual factors and message factors, which left studies in medium factors in a surprising scarcity (McGuire,1969). Such a scarcity was also found in the social media advertising literature, as discussed earlier. Wright (1974) attributed the lack of medium research to the difficulty of identifying consistent dimensions for comparing mediums. He argued that medium such as television, magazines, radio, etc. are different in many dimensions simultaneously, and even many of the dimensions are not necessarily comparable due to the variation. The same difficulty also exists for social media platform comparison due to the great variety of social media platforms, such as content community, social networking sites, online game community, etc. Stewart and Ward (1994) and Steward et al. (2002) suggested an alternative approach to understand the medium differences. Specifically, they argued that there should be a switching of the focus in medium studies, i.e., from examining medium characteristics to studying the ways individuals interact with and act upon the medium. Specifically, with today’s rapid evolution and change of social media landscape, it is difficult to truly understand the medium effects without considering how the social media is being used by the users. Therefore, taking this alternative approach in understanding medium, a theoretical construct that specifically focused on the differences across different social media platforms based on individuals’ prior use of the medium was developed and examined in the current research project. The purpose of the current research project has threefold. First, the researcher intends to 3 provide a constructive review of extant empirical studies about social media advertising literature through analyzing how prior scholars have approached this topic. Specifically, following the classic communication model, the researcher reviewed 133 empirical studies published in the top 10 advertising and marketing journals during 2006 – 2016. The findings of the literature review revealed the lack of medium-factor studies in understanding social media advertising. Secondly, the researcher introduces the concept “enacted affordances of social media.” It’s a concept that was developed to understand what each social media platform can provide based on individuals’ actual usage behaviors. Specifically, a taxonomy of the enacted affordances of social media was proposed based on two dimensions: 1) nature of the content (relational vs. non-relational) and 2) action on the content (consuming, contributing and creating). Moreover, the current research project applied this new concept in an empirical setting to investigate its relationship with consumers’ responses to social media advertising. A series of hypotheses were proposed in the current study to understand how specific enacted affordances could relate to consumers’ acceptance and avoidance of social media advertising, as well as what kind of social media platforms would be considered as more advertising-friendly in consumers’ opinion. To our knowledge, the concept enacted affordance of social media is the first theoretical concept that was developed to understand the medium effect in social media. And the current project also provides the very first empirical study that examines the applicability of such concept in advertising studies. The newly developed concept itself would be beneficial to both scholars and practitioners in communication field to further distinguish the social media platforms from individuals’ usage behavior. Moreover, the taxonomy of the enacted affordances of social media could serve as a practical guideline for communication professionals and media planners to make proper media choices and tailor communication strategies when intending to achieve effective 4 communication outcomes. On the other hand, the empirical part of the current research project specifically focused on consumers’ social media advertising acceptance and avoidance, the findings of this project could help advertising practitioners understand how platforms differ in terms of consumers’ advertising acceptance and avoidance, as well as what kind of enacted affordance of social media would have a stronger impact on such acceptance/avoidance behavior. Moreover, the findings could also benefit social media technology companies that offer advertisements, such as Facebook, Twitter, Instagram, to understand how to increase users’ advertising acceptance through adjusting the enacted affordance of the social media platform. The remainder of this research paper proceeds as follows. The researcher first provided evidence for the lack of medium studies in social media advertising through a literature review of 133 empirical studies published in eleven top advertising and marketing peer-reviewed journals during 2006 – 2016. Following the review, the research introduced the concept of enacted affordances of social media as the new theoretical lens to study medium differences across different social media platforms. Elaboration on the taxonomy of the six types of enacted affordances of social media is also provided. Lastly, applying the concept in social media advertising, the researcher empirically examined the relationship between the enacted affordances of social media and consumers’ acceptance and avoidance of social media advertising. 5 CHAPTER 2 REVIEW OF SOCIAL MEDIA ADVERTISING Topics of advertising-related studies in social media have evolved dramatically over the years, from early forms of electronic word-of-mouth (e.g., Smith et al. 2007; Dwyer, 2007; Prendergast et al. 2010), brand community (e.g., Casaló et al. 2008; Schau et al. 2009), viral advertising (e.g., Hinz et al. 2011; Van Noort et al. 2012) and consumer engagement (e.g., Muntinga et al. 2011; Tsai & Men 2013; Holleeek et al. 2013), to today’s targeted sponsored content and consumer conversion (e.g., Keyzer et al. 2015; Van Reijmersdal et al. 2016). Due to social media’s fast-evolving landscape and diversification, new formats of social media advertising were growing and transforming at a fast speed. Such continuous change of social media platforms and the advertising products offered in these platforms make the definition of ‘social media’ and ‘social media advertising’ difficult to reach its full comprehensiveness. In the current research project, we adopt Kaplan and Haenlein’s (2010) broad definition of social media, which was defined as ‘web based applications that built on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of user generated content.’ Built upon this understanding of social media, the current study thus proposes the definition of social media advertising as “any persuasive content distributed through social media platforms for consumer engagement or consumer conversion.” Moreover, the categories of social media advertising range from owned social media advertising (e.g., brand page updates), paid social media advertising (e.g., in-feed native advertisement) to earned social media advertising (e.g., shared viral advertisement). A more detailed discussion regarding the concept of social media advertising will be provided in a later chapter. 6 Therefore, using this broad perception of social media advertising, the researcher reviewed 10 top journals in advertising and marketing fields during 2006 – 2016 and found 159 published studies that falls under the scope of social media advertising, i.e., having “social media” and “advertising” in either the title of the article, abstract of the article or the keywords of the article. The purpose of the current literature review is to provide a holistic overview regarding the scientific approaches scholars have taken in understanding social media advertising. Specifically, the published articles were analyzed through a classic communication process (Lasswell 1948; Stern 1994) with four main factors being identified as majorly investigated input variables: individual factors (to whom the message is received), source factors (from whom the message is sent), message factors (what being communicated) and medium factors (through which channel message is communicated). Excluding articles that were conceptual papers or uses simulation modeling, a total of 133 empirical studies of social media advertising were analyzed in the final process. Some empirical studies covered more than one kind of input variables, e.g., studies investigating the interaction effects of individual factors and message factors. Therefore, some of the studies could be counted for more than once, if it involved more than one kind of input variables. Table 1 presents the statistical details regarding the number of articles published in the past 10 years in the top ten advertising and marketing journals. Table 2 presents the overall thematic categorizations for each type of factors. As indicated in Table 1 and Table 2, majority (67%) of extant literature in social media advertising focused on the individual factors, such as individuals’ motivation in participating brand community or engaging with advertisements in social media (Muk & Chung 2014; Kwon et al. 2014), 2) individual differences in various aspects such as their content creating behavior (Morrison et al. 2013) or their gender differences (Hoy & Milne 2010; Do et al. 2014; Kim and 7 Yoon 2014), 3) individuals’ identification or interaction with brand (Yeh & Choi 2011; Do et al. 2014; Shan and King 2015), as well as their concerns over privacy issues (Jeong and Coyle 2014) and advertising avoidance in social media (Kelly et al. 2010). It is not surprising to see such a huge effort being focused on the individual factors since the users are the core driving power of the sustainability of any media platforms. Meanwhile, a fair amount of studies also investigated the impact of message factors (34%), such as characteristics of advertising messages (e.g., De Vries et al. 2012); and source factors (26%), such as perceived similarity (e.g., Thompson & Malaviya 2013; Paek et al., 2015), perceived tie-strength (e.g., Dubois et al. 2016), on various consumer responses to social media advertising. However, only a few studies (10%) focused on the medium effect in social media advertising. Since the current research project takes the medium approach to understanding consumer responses to social media advertising, the following part will review extant literature that involved medium factors in depth. Medium-factor Studies in Social Media Advertising Studies relating to medium effects in social media advertising had the least attention among all in the literature. A total of 13 studies were published in the reviewed top 10 advertising and marketing journals during 2006 - 2016. Although the number of publications was a small number as compared to the other types of factors, several streams of medium studies were still identified in the reviewing process. Specifically, earlier scholars compared social media with traditional media and examined which kind of media type drives better advertising effectiveness (e.g., Trusov et al. 2009). As the growing variation of social media platforms, scholars also started to notice the different roles various social media platforms play in advertising (e.g., Smith et al. 2012). Moreover, scholars also explored how individuals’ perception or relationship with the medium influences individuals’ response to messages (e.g., Wu 2016) and conducted macro analyses of 8 different social ‘medium’ (e.g., Seraj 2012). The following section elaborates each of research stream in detail. Comparison of Social Media with Traditional Media When a new type of media is introduced into the market, one of the main questions to be answered is ‘how is it different?’ Therefore, earlier studies have majorly investigated how the persuasion outcomes differed between social media and traditional media. Trusov et al. (2009) compared the effect of word-of-mouth marketing on membership growth between online social networking site and traditional marketing vehicles. They found that the social networking site had a substantially larger effect than the traditional kind of marketing vehicles. Pfeiffer & Zinnbauer (2010) practiced a marketing mix modeling to compare the effectiveness of online channels and traditional communication channels. Overall, they found that TV had the most influential impact, yet the social networking site offers great opportunities for driving website traffics from electronic word-of-mouth. Moreover, Mabry & Porter (2010) compared between social networking sites and official website of a movie to find the most effective mean in increasing consumers’ intent to watch. Although the findings indicated that the official website was more effective than the social networking promotion page, the authors suggested that the most effective approach should be a combination of both approaches. Such a suggestion regarding the combination effect between traditional media and social media was then further studied by later scholars. For example, Spotts et al. (2014) found a reciprocal relationship between television and social media. Specifically, they suggested that television can significantly amplify social media conversations about the advertised brand and both platforms together enhanced brand engagement. Similarly, Nagy & Midha (2014) also suggested that the interaction between television and social media is highly complementary in building consumer engagement. 9 Comparison of Different Social Media As the diversification and growth of social media platforms, several studies have also been conducted to understand how social media platforms differ from each other, as well as how such differences influence consumer behavior. Smith et al. (2012) first compared the differences in brand-related user-generated content (UGC) across Twitter, Facebook, and YouTube. Through a content analysis of 600 UGC posts, they categorized six types of UGC content, namely, promotional self-presentation, brand centrality, marketer-directed communication, response to online marketer action, factually information about the brand, and brand sentiment. They’ve also concluded with how the three social media platforms differed across these six categories. Moreover, Van-Tien Dao et al. (2014) took the approach of considering social media type as moderator in influencing the relationship between the antecedents of social media advertising value and consumers’ purchase intention. The findings showed significant moderation effect. Comparing with content community website (e.g., youtube), the effects of advertising informativeness and entertainment on perceived advertising value is stronger than those on social networking sites (e.g., Facebook). In other words, YouTube differs from Facebook regarding how the perceived advertising value can be influenced by the advertisement beliefs presented in the social media advertisement. Logan (2014) also further suggested that individuals following brands on different social media platforms could be seeking different user gratifications. Individuals’ Relationship with Medium Scholars have also taken the approach of looking at the medium difference from individuals’ relationship with the medium. Three major studies were found in the extant literature that discussed such relationship. Lee & Ahn (2013) examined the role of perceived medium credibility and its interaction with perceived advertisers’ motive on consumers’ intention to participate facebook 10 brand page. Results showed that when the credibility of the social media platform is high, consumers were more likely to engage in brand activities. VanMeter et al. (2015) proposed the concept of consumers’ attachment to social media (ASM) and developed a scale for the measurement. They defined consumers’ attachment to the social media as the strength of a bond between a person and social media. Their empirical study showed that one’s attachment to social media has a positive influence on one’s C2C advocacy and C2B supportive behaviors. From a similar vein, Wu (2016) investigated the role of media engagement on the perceived value and acceptance of advertising within mobile social networks. The findings showed that one’s media engagement could positively influence perceived advertising value and consumers’ acceptance of social media advertisement. Macro Analyses of Social Medium Besides the micro-investigation of medium effects in social media advertising, scholars have also taken a macro-approach in understanding the social media as a medium itself. For example, in the early days of social media, Seraj (2012) explored the main characteristics of an online community that can deliver value to its consumers and instigate engagement. The study revealed three specific online community characteristics that create value for the members, namely: goal driven and quality content (intellectual value), interactive environment for building relationships (social value) and self-governed community culture consistent with its principles (cultural value). A more advanced metrics of social media was developed later by Peters et al. (2013), in which they discussed what elements would constitute a social medium. In total, they categorized four major elements of social media metrics, namely, ‘motives,’ ‘content,’ ‘network structure,’ and ‘social roles & interactions’. ‘Motive’ refers to the motivations that drive individuals’ action in social media, ‘Content’ refers to the content being carried along the social 11 ties in social media, ‘Network Structure’ refers to the underlying connection structure in the social media, while the ‘Social Roles and Interaction’ refers to the continuously mediated relationship between the individual and the surrounding network, which is processed through social interactions. A guideline for social media metrics and dashboards was also proposed in the paper. Overall, Peters et al. (2013) provided a constructive framework for both practitioners and scholars to understand social media from a more macro approach. Conclusion Similar to the scarcity of medium-effect studies in traditional communication areas, there is also a lot of space for the growth of more empirical studies of medium effects in social media advertising realm. Although extant literature has covered various aspects of medium effects in social media advertising, such as comparing traditional media with social media regarding wordof-mouth effectiveness (Trusov et al. 2009), or comparing different social media platforms on the content characteristic (e.g., Smith et al. 2012), none of the extant studies drew the connections between various medium differences and advertising effectiveness. Wright (1974) attributed the lack of research regarding comparative medium-effect studies partially to the difficulty of identifying the dimensions that can differentiate one medium platform from another. In other words, the lack of comparative medium-effect studies is partially due to the lack of a consistent ruler’ that can measure the commonly shared dimensions across mediums. The researcher acknowledges the difficulty of drawing consistent commonly shared dimensions if only looking at physically existing features across various platforms, since many of those features were not comparable due to its variations. However, if we were to consider the media differences from individuals’ perception regarding how the platform could offer and how the platform could be used, then we could draw clear lines in between the differences of the media platforms. 12 Therefore, in the following section, the researcher will introduce a new concept tailored to the need of understanding social media platform differences. Termed as ‘enacted affordances of social media,’ the concept is defined as the acted-out action possibilities by a user on a social media platform to achieve particular goals in a specific context. Built upon two dimensions, namely the nature of the content and the action on the content, a taxonomy of six types of enacted affordances are proposed to serve as the “ruler” in understanding how social media platforms differ from each other on each of the dimensions. Through this new theoretical lens, scholars can not only understand the strength of each social media platforms, but also draw connections between the different types of enacted affordances of social media and consumers’ responses to various advertising messages on different social media platforms. It advances the current social media advertising literature by opening up new angles for scholars to understand and investigate the social media advertising effectiveness. The following chapters will elaborate on the rationalization and application of this ‘ruler’ in detail. 13 CHAPTER 3 A HYBRID APPROACH OF MEDIUAM AND INDIVIDUAL: ENACTED AFFORDANCES OF SOCIAL MEDIA The concept of affordance was first introduced through the affordance theory by Gibson (1979), which discusses the dynamic relationship between an individual/organism and its environment or another object. The fundamental meaning of affordance refers to ‘all the possible offerings the environment/object can provide to an individual.’ However, different from the concept of ‘affordance,’ Norman (1988) introduced the concept of ‘perceived affordance,’ which he distinguished by arguing that it is the perceived affordances of an object that determine the usability of an object, not all the possible affordances of an object. Bærentsen and Trettvik (2002) further extended the understanding of perceived affordance through introducing the activity theory approach which discussed that actual interaction between the subjects and objects are important factors in determining the perceived affordances of the object. Following up on the Bærentsen and Trettvik’s (2002) approach of activity theory, the current study thus proposed the concept of enacted affordance as the acted-out actions possibilities by an individual on the environment/object. Such an enacted affordance concept is different from the perceived affordance proposed by Norman (1988) for the affordance has to be actually activated by the user, rather than simply perceived by the user. In other words, if an individual not only perceives the affordance of an object but also performs the action to realize the affordance of the object, then we consider such a realization of the perceived affordance as the enacted affordance. Therefore, applying this series of development of ‘affordances’ to social media, the ‘enacted affordances of social media’ can be understood as ‘all the actions an individual enacted 14 among the action possibilities offered/provided by the social media’. Based on the literature review of individuals’ actual action/utilization of social media, two dimensions are thus identified to understand the enacted affordances of social media. In the following part, the researcher first elaborates the theoretical foundation and development of the concept, and then introduces the taxonomy of the enacted affordances in social media. Affordances, Perceived Affordances and Enacted Affordances Rooted in ecological psychology, Gibson (1979) proposed the concept of affordance and stated that the affordance of the environment is ‘what it offers the animal, what it provides or furnishes, either good or ill’ (Gibson, 1979, p. 127). Therefore, affordance of the environment is realized through the interaction between the organism and the environment. It addresses the complementary relationship between the abilities of an organism and the features of the environment (Chemero, 2003). Instead of static properties, the affordances of the environment are made up of the dynamic relational attributes between organisms and the environment (Zhao et al., 2013). The existence of an affordance has three fundamental properties, as concluded by McGrenere and Ho (2000) that an affordance exists relative to the action capabilities of a particular actor; the existence of an affordance is independent of the actor’s ability to perceive it; and an affordance does not change as the needs and goals of the actor change. Therefore, the majority of the Gibsonian psychologists believe that the affordance of an object is not influenced by an individual’s experience, knowledge, or ability to perceive; rather, it is determined by the action capability by an actor in the environment. However, in his book “The Psychology of Everyday Things,” Norman (1988) presented a different viewpoint. Specifically, he stated that affordances result from the experience, knowledge, 15 or culture of an actor, which can influence the individual’s understanding of how an object can be used. Specifically, clues to the operation of things exist due to one’s previous experience, knowledge or culture, and through which it suggests the range of action possibilities (Norman, 1988). For example, a chair can afford sitting, it can also provide support on which actors can put objects on. A bottle of water can afford water drinking, but it can also be used for exercising as weight lifting. Therefore, affordances of an object are often time in association with one’s mental perception of how to use the object. In his recent studies, Norman (1999, 2008) discussed the difference between perceived affordances and real affordances. He emphasized that it is the perceived affordances that determine the usability of an object, not the real affordances. In other words, Norman regarded the affordances of an object as being determined by how an individual perceives the object shall be used, rather than how an object can actually be used or how many ways it can be used. From his viewpoint, it is important for the organism to pick up information regarding the uses of the objects. Such an approach highlights the importance of one’s understanding and expectations of interaction with the object in design (McGrenere and Ho, 2000). From the activity theory approach, Bærentsen and Trettvik (2002) further discussed the importance of actual concrete interaction between the actor and the artifacts in understanding affordances. Their viewpoints upon affordances propose that an affordance can only be perceived based on the perception of events --- actual interaction between subjects and objects. Therefore, activity is the key link that connects organism and its surroundings. As for an affordance to exist, such an observation of interaction has to be acquired in forming the perceived affordances. Bærentsen and Trettvik’s (2002) activity theory approach extends the notion of observing “activity” in understanding affordances. Similarly, Gibson and Pick (2003) also discussed the importance of 16 “learning” in understanding perceived affordance. They argued that actors discover affordances of objects through perceptual learning because affordances do not automatically present themselves to actors. In general, the key difference between Gibson and Norman’s understanding of affordances lies at if the “affordances” of an object can be perceived or not, i.e., in Gibson’s definition, an affordance exist regardless of an individuals’ capability to see it or not, it exists objectively; while, in Norman’s definition, an affordance exists when an individual, based on his/her previous experience, education, culture factors, or as Bærentsen and Trettvik (2002) argued – the observation of such interaction, can perceive such action possibilities with the object, which is subjectively existing. This researcher agrees with the viewpoint of “perceived affordance” because in the social media realm, various individuals, under various contexts, can use social media platforms for different reasons. The researcher valued the activity theory approach more as Bærentsen and Trettvik (2002) shed heavier attention on the actual interaction. The affordances of a social media platform are not merely properties of the platform itself, nor the users’ characteristics, but the intersection between these two. How an individual uses the social media platform, thus, determines the affordances of the social media platform. Such affordances could be a routine for many of the consistent users on a social media platform, but it could also be dynamic and flexible among different user groups, under different contextual situations. Thus, the researcher adopted the word “enacted affordance” in the current study for it extends the “perceived affordances” to the further notion of actual interaction. The following section would elaborate the adaption of ‘enacted affordance’ in social media in further details. 17 Enacted Affordances of Social Media To adapt the concept of enacted affordances in social media, the researcher considered the specific social medium as the object, while the individual user of the medium as the actor. The actual affordances of the social media include all the possible actions one can conduct on the medium, while the enacted affordances of social media include only the actions that one conducted on the medium. In today’s social media realm, although all the platforms facilitate social interaction, they vary in terms of the extent to which “social interaction” is being focused. Although the actual affordances of social interaction across various social media platforms might be the same, the enacted affordances of social interaction on different social media platforms for each individual could be different. Therefore, enacted affordance of social media is defined as the acted-out action possibilities for a user to conduct on a social media platform to achieve the goals in a particular context. Specifically, such affordance is a dynamic relationship between the medium itself (e.g., user interface design) and the individual’s characteristics (e.g., personal network structure, previous experience, knowledge, or personality), which could also be altered by the contextual environment (e.g., alternative choices, situational factors). Media Characteristics and Social Media Use As a designed object, the social media platform itself has great influence in guiding how individuals could use the platform. For example, Keenan and Shiri (2009) compared sociability and social interaction on four social networking websites, namely, Facebook, MySpace, Twitter, and LinkedIn. They explored the ways that these websites encouraged sociability and analyzed the design practices of each social media platforms. Their results revealed various interaction options on these websites, including media sharing, blogging, public discussions, events, and a host of 18 other features. After comparison, they found that large-scale social media sites such as Facebook and MySpace have a multitude of features, while small-scale social media sites such as LinkedIn and Twitter are more niche-focused with fewer features. Moreover, they argued that the sociability of social media is largely determined by the design of the website. Examples given in their study emphasized the limited SMS-length of Twitter, which resulted in encouraging “quick updates,” and the public online persona of MySpace, which resulted in a higher level of publicity but lowprivacy features. Along with the above-mentioned studies, scholars have also investigated specific features on social media that can influence interactive use. Viswanath et al. (2009) analyzed real data from Facebook with a focus on network relationship and user interaction. Their findings showed that Facebook’s birthday reminder feature contributed to 54% of the interactions between infrequently interacting user pairs. Additionally, Zhang et al. (2013) studied the Pinterest-style infinite scroll layouts and found that such dynamic grid layouts and endless scrolling increased users’ average time spent on the website, as well as their appreciation of others’ sharing. But it also decreases users’ original sharing intention. Individual Characteristics and Social Media Use Since enacted affordances of social media are one’s acted-out actions of the action possibilities a social media platform provide, individual characteristics, thus, are also involved in forming individuals’ understanding of the social media affordances. Individual characteristics, such as personality (Schrammel et al., 2009; Correa et al., 2010; Hughes et al., 2012; Seidman, 2013) or previous experience (Zhang et al., 2013), could affect social media use. Specifically, Correa et al. (2010) found that one’s extroversion and openness to experiences are positively associated with one’s social media use frequency. Schrammel et al.’s (2009) also 19 found a positive relationship between one’s openness and his or her number of friends on social media. Seidman (2013) found that individuals high in agreeableness are more likely to share selfexpression content on their Facebook Timeline, while highly conscientious individuals are less likely to post information about themselves since they are more cautious about their self-images. Moreover, introverted users were found to be more likely to consume their social networks’ News Feed to catch up with friends (Moore and McElroy, 2012). In addition to these findings, Hughes et al.’s (2012) study compared users’ individual characteristics and their usage of two social media platforms: Facebook and Twitter. Results showed that people who are more gregarious and sociable are more likely to use Facebook, while people with a high need for cognition are more likely to use Twitter for information seeking. The above discussion built on extant literature explains the two major components (media and individual) of enacted affordances of social media in influencing one’s social media use. Yet, the “goal” element of enacted affordances of social media have not been addressed. Thus, the following section will introduce the role of a “goal” in influencing enacted affordances of social media. Specifically, this study argues that individuals’ goals of using a social media platform interact with their individual characteristics and the media characteristics in determining how they consider the social medium can be used. Therefore, the typology of enacted affordances of social media is the product of analyzing media attributes, individual characteristics, and goals of using social media. Typology of Enacted Affordances of Social Media Just as Gibson (1979) proposed the three fundamental properties of an affordance that an affordance exists relative to the action capabilities of a particular actor; the existence of an affordance is independent of the actor’s ability to perceive it, and an affordance does not change 20 as the needs and goals of the actor change, the current study also proposes the three fundamental properties of an enacted affordance: an enacted affordance exists relative to the action capabilities of a particular actor; the existence of an enacted affordance is dependent on the actor’s ability to act it; and an enacted affordance changes as the needs and goals of the actor change. The first two properties have already been addressed in the previous text; next, the researcher elaborates on the role of a goal in understanding enacted affordances of social media. Specifically, existed literature had found that one’s social media activities are usually goal-driven (Hoffman and Novak, 2012). Therefore, the following section will review extant literature on individuals’ goals of using social media and the related activities one would conduct to achieve the goals. The relationship between such activities and the enacted affordances of social media is also discussed. Social vs. Non-Social Goals of Social Media Use Inherent in its nature, social media is a tool for people to socialize with each other. When it was created, such computer-mediated communication among individuals broke the boundaries of physical face-to-face interaction and was found to have a positive impact on interpersonal relationships (Tidwell and Walther, 2002). The majority of previous studies that investigated individuals’ motivations for using social media found the characteristic “social” to be a key indicator in motivating consumers’ usage of social media (Donath, 2007; Ellison et al., 2007; boyd and Ellison, 2007). In particular, studies of social networking sites found these sites to be a great platform for individuals to present themselves, establish and maintain social relationships, and articulate their own social networks (Ellison, 2007). However, Hoffman and Novak (2011) argued that just because social media is named “social,” not everything that happens on social media platforms involves a high level of social interaction. Such an argument is in line with the earlier statement regarding perceived affordances; 21 that is, although all social media platforms facilitate “social interaction,” the extent to which the social interaction is focused can vary. Such variation could be determined by the medium itself. Kaplan and Haenlein (2010) categorized social media types into collaborative projects, blogs, content communities, social networking sites, a virtual social world, and a virtual game world, which indicated various focuses. And the previous discussion regarding the role of interface design (e.g., Facebook’s birthday reminder feature) in influencing one’s social interaction (e.g., between an infrequently interacting pair) also supports this statement. Moreover, the variation could also be associated with one’s individual characteristics, for example in the finding that people who are extroverted, highly agreeable, and very open are more likely to socialize with others (Schrammel et al., 2009; Selfhout et al., 2010; Golbeck et al., 2011). In order to explore the other individuals’ motivation when using social media, previous scholars have also found that social media could be used for information seeking (e.g., Park et al., 2009; Whiting and Williams, 2013; Lee and Ma, 2012), self-expression (e.g., Shao, 2009; Courtois et al., 2009; Kaplan and Haenlein, 2010); passing time or killing boredom (e.g., Quan-Haase and Young, 2010; Whiting and Williams, 2013), and entertainment (e.g., Park et al., 2009). Specifically, Hoffman and Novak (2012) conducted a literature review on motivations studies of social media use and proposed 22 social media goal types, which were further summarized into social and non-social goals. Social goals refer to the intent to get connected with friends and family, meet new people, or reconnect with old friends, and to express oneself through sharing content. Non-social goals refer to the intent to learn about new things or trending events, to get entertainment through music or video clips, to find deals, or to obtain product information. Hoffman et al. (2016) further elaborated on the social and non-social activities on social media. They termed “social behaviors” as activities that involve direct and active interaction with others 22 (e.g., finding or contacting friends, sending messages, and commenting on friends’ posts). “Nonsocial behaviors” were defined as activities that involve indirect and passive interaction with others (e.g., reading news, getting information, finding product reviews, and checking out trending topics). However, after reviewing the definitions of “social” and “non-social” activities on social media (Hoffman et al., 2016), the researcher found that there were two dimensions involved in one’s social media activity: nature of the content and actions on the content. Specifically, the researcher argues that “social behaviors” on social media do not have to be always active; simply consuming friends’ updates is a form of social behaviors. On the other hand, “non-social behaviors” on social media do not have to be always passive, sometimes users can also create information that is not relevant to socialization purpose on social media, such as product reviews, etc. The researcher acknowledges the existence of social and non-social goal when individuals use social media platforms, yet, when it comes to individuals’ social media use, it should be viewed from the dimensions of content and means. Specifically, the content can be categorized into “relational” content and “non-relational” content. Relational content is often used for satisfying one’s social goal. It refers to content that is person-focused, made for initiating, maintaining or building the personal relationship on social media. Examples include activities or events of one’s personal life shared on Facebook. Non-relational content, on the other hand, is often used for satisfying one’s non-social goals. It refers to content that is information-focused, either entertaining or useful. It’s not for the interpersonal relationship, but for one’s interests. Examples include commercial information, news articles, or funny videos. “Consuming”, “Contributing” and “Creating” on Social Media 23 Shao (2009) proposed three types of activities that people perform online, i.e., consumption of information and entertainment, participation in social interaction and community development, and production of self-expression and self-actualization. Similar to Hoffman and Novak’s (2016) mixing of content and means of communication, Shao’s categorization introduced context for each means of communication. For example, consumption of information and entertainment could also be consumption of relational content on social media; while production of self-expression and selfactualization could also be the production of useful or entertaining information. Therefore, the researcher, again, argues that a clear separation should be made between the nature of the content and the actions on the content. As previously indicated, the researcher has identified two kinds of content nature: relational vs. non-relational. Therefore, the following section will elaborate on the actions on the content. Muntinga et al. (2011) categorized consumers’ brand-related activities on social media, including consuming, contributing, and creating. Consuming refers to consumers’ consumption of brand information or others’ reviews about brands on social media. Contributing refers to consumers’ active participation with brands on social media, such as liking a brand’s fan page or commenting on a post. Creating—the highest level of online-related activeness—refers to consumers’ active production or publishing of brand-related content to which other consumers can contribute. This kind of consumer activity often relates to user-generated content on social media. Such a distinction of consumers’ activities on social media is believed as a better choice that can also be applied to individuals’ general usage of social media. Therefore, the current study adopts Muntinga et al.’s (2011) understanding of consumer interaction on social media and suggests that there are three means of communication when it comes to general uses of social media: consuming, contributing and creating. Consuming 24 addresses individuals’ acquisition of various content on social media; it could be either content about their social connections or content of one’s interest. Contributing refers to individuals’ response to others’ initiated content, it could be either “liking,” “commenting” or other equivalent behaviors on social media. Creating, on the other hand, refers to one’s user-initiated content on social media; it could be content that is purely user-generated or, it could be content shared among users. Enacted affordances of social media, as a means to understand social media characteristics based on individuals’ uses of the social media, thus, is categorized from two dimensions: nature of the content (relational vs. non-relational) and actions on the content (consumption, contribution, and creation). Table 3 shows the categorization in details with explanations for each scenario. As shown in Table 3, six types of enacted affordances of social media are proposed, namely, relational content consumption, relational content contribution, and relational content creation, non-relational content consumption, non-relational content contribution and non-relational content creation. By definition, enacted affordance of relational content consumption refers to one’s evaluation of how the social media platform could be used to consume relational content, such as posts about one’s social connections’ life; enacted affordance of relational content contribution refers to one’s evaluation of how the social media platform could be used to contribute relational content, such as providing feedbacks through “liking” or “commenting” on others’ posts about interpersonal relationships; enacted affordance of relational content creation refers to one’s evaluation of how the social media platform could be used to create relational content, such as initiating posts about one’s personal life; enacted affordance of non-relational content consumption refers to one’s evaluation of how the social media platform could be used to consume nonrelational content (useful/entertaining), such as reading news articles or watching funny video clips; 25 enacted affordance of non-relational content contribution refers to one’s evaluation of how the social media platform could be used to provide feedback on non-relational content, such as “liking” or “commenting” on others’ posts about non-relational content, such as product information, news articles and etc.; enacted affordance of non-relational content creation refers to one’s evaluation of how the social media platform could be used to contribute non-relational content, such as creating online product reviews or sharing useful or entertaining information, such as news articles and funny videos. These six types of enacted affordances of social media, although independent from each other, together it frames one’s overall understanding of his/her uses of the evaluated social media medium, which can be presented through a hexagonal shape, with each enacted affordance given a value from 1-7. The operationalization of the concept was also developed in the researchers’ earlier prelim project. The final adopted measurement included 18 items for measuring the six types of enacted affordances of social media. Table 4 presents the scale measurement in detail. Given the lack of medium studies in extant communication and social media research, the current concept has its significant contribution regarding introducing a new theoretical perspective to understand various social media platforms. Specifically, as suggested by Steward and Ward (1994) that the focus of today’s media studies should have a change of focus from the stimulus -- medium--- to the individuals who are interacting with it, the current concept focused on the relationship between the user and the medium regarding how an individual perceive what the medium can afford. Thus, it advances as a concept to define individual-based medium characteristics. Moreover, as suggested by Wright (1974), the lack of medium studies in communication could partially be attributed to the difficulty of identifying the dimensions to consistently 26 differentiate one medium from another. The current proposed concept solves this problem by providing the taxonomy of the six types of enacted affordances of social media. The operationalization of this concept gives scholars a consistent ruler to compare the differences across various social media platforms. With such a consistent comparison, communication scholars can draw connections between variously perceived affordances of social medium to the communication outcomes to understand why specific social media platforms have a better persuasion results than others. The following chapter will present a study focusing on the implication of this concept in social media advertising. 27 CHAPTER 4 ENACTED AFFORDANCES AND CONSUMERS’ RESPONSE TO SOCIAL MEDIA ADVERTISING The purpose of this chapter is to explore the implication of enacted affordances of social media in the field of social media advertising. Begin with a discussion of today’s social media advertising product offerings, the researcher defines social media advertising as “any paid persuasive content distributed through social media platforms for consumer engagement or consumer conversion”. Following this definition, the current study specifically investigated how different enacted affordances of social media could contribute to consumers’ response toward paid social media advertising, since the paid advertisement in social media provides consumers the least control over receiving the advertisement. The findings of the current study would be beneficial for digital media planner and social media technology companies to understand which kind of social media platforms would be more advertisement-friendly from consumers’ perception, and which kind(s) of perceived social media affordance would be most relevant in driving consumer response. Scholars in advertising field could also gain new insights about the comparative medium-effect in social media advertising. Social Media Advertising: A Paid Format of Marketing Communication The term “advertising” was once defined as “any paid-for communication intended to inform and/or influence one or more people.” (Bullmore, 1975). Based on this definition, the “paid” component plays a key role in advertising. However, with the development of computer-mediated communication on the Internet, especially with the rise of social media, more forms of advertisement started to draw attention, such as user-generated content and online viral 28 advertisement through electronic word-of-mouth. Therefore, a latest revised definition was proposed, Bullmore (2016) revised his definition of advertising as “any communication, usually paid-for, specifically intended to inform and/or influence one or more people”. Such a revision although provides a broader perception of the concept of advertising, it also highlights the importance of “paid” element in the advertising industry. Adopting this perception of advertising, the researcher discusses the concept of social media advertising in today’s growing scope of marketing communication in social media. Hurrle & Postatny (2015) grouped social media into paid, owned and earned media. Adapting this categorization and reflect upon the historical development of social media marketing, a clear pattern of the marketing communication development in social media is revealed. In the early days, when social media were first becoming popular, owned content and earned content were the focus of practices and researchers. Through owned social media for marketing communication, brands can create brand pages in social media to directly post brand content on its page, as well as interacting with its followers. While, owned social media communication occurs when individuals engage with brand content through “liking,” “sharing,” “commenting” or other forms of responses (varied due to the social media platform differences). Scholar investigated such earned content exposure in social media through investigating topics such as electronic wordof-mouth (e.g., Smith et al. 2007; Dwyer, 2007; Prendergast et al. 2010), viral advertising (e.g., Hinz et al. 2011; Van Noort et al. 2012), and consumer engagement (e.g., Muntinga et al. 2011; Tsai & Men 2013; Holleeek et al. 2013). Along with the development of social media, standardized advertising offerings were gradually developed across social media platforms. Specifically, with the introduction of “sponsored content” on leading social media platforms starting early 2011, the spending on social 29 media advertisement started to rise at a rocketing speed. It was shown that from 2012 to 2016, the spending of social media advertising in the United States grew from 4.3 billion to 11.7 billion dollars, and it is predicted to reach 15 billion dollars by 2018 (BIA/Kelsey on Statista, 2017). Scholarly publications started to discuss the effectiveness of such targeted advertisement (e.g., Keyzer et al. 2015; Van Reijmersdal et al. 2016). The paid advertising products offered in social media platforms, thus, become the major power in today’s social media advertising industry. One of the main differences in between the owned earned and paid marketing communication is consumers’ control over receiving the advertisements. With owned media, consumers have the highest control regarding what kind of advertisement they see since they can decide which brand to follow in social media to receive brand content. While with earned media, consumers have a lower control as compared to the owned media, since consumers can’t control what kind of advertising message their social networks share. However, on the other hand, since earned media comes from consumers’ social connections, consumers can control from whom they can receive messages. For those whom they don’t want to receive messages, consumers can block information from them. However, with paid media, which is addressed as the social media advertising, consumers don’t have much control regarding from whom and what kind of advertisement they will receive. In other words, consumers can only passively accept advertising messages, unless they install ad blocker. Therefore, in the current study, the researcher focused on this growing power of paid media in social media communication --- social media advertising. Consumer Acceptance and Avoidance of Social Media Advertising Consumer Acceptance of Social Media Advertising The term “acceptance” was known and examined in communication research majorly due to the introduction of technology acceptance model (Davis 1989). Based on its original elaboration 30 in the field of information system, scholars studying one’s “acceptance” of technology focused on two main factors: “one’s attitude toward the technology” and “one’s behavioral intention to use the technology” (e.g., Jackson et al. 1997, Moon & Kim 2001). Adopting this model, scholars studying advertising in a new technology environment also measured individuals’ acceptance through “one’s attitude toward advertising” and “one’s behavioral intention to accept the advertising in the technology”, e.g., mobile advertising (Tsang et al. 2004, Zhang & Mao 2008), mobile social networking advertising (Wu 2016) etc. However, the term attitude was defined as “a learned predisposition of human beings” by Fishbein (1967) and based on which, “an individual would respond to an object (or an idea) or a number of things (or opinions)”. In other words, individuals’ behavioral intention to accept the technology is the consequence of one’s attitude toward the technology. Therefore, the current study argues that the nature of consumers’ acceptance of social media advertising should focus more on individuals’ behavioral outcomes. Moreover, the researcher also notices that accepting paid advertising in social media is very much different from accepting a technology in information systems. In the information system domain, one’s acceptance of technology is under one’s own control, such as one’s acceptance of mobile commerce is made based on one’s own will in using the mobile commerce or not. While in the social media advertising domain, the advertisements are forcefully exposed to the consumers during their use of social media platforms, which left them little control regarding whether they will receive advertisements or not, from whom they will receive advertisement and what content will be presented in their social media platforms. In other words, the word “acceptance” in social media advertising is more of a passive behavior than an active behavior. Such argument is also supported by the definition of Oxford Dictionary, in which they defined the word “acceptance” as 31 “the action of consenting to receive or undertake something offered”. Therefore, based on the above discussion, the current study considers consumers’ acceptance of paid social media advertisement as merely “consumers’ willingness to receive paid advertisement in the social media.” Consumer Avoidance of Social Media Advertising Advertising avoidance was defined by Speck and Elliott (1997, p.61) as “all actions by media users that differentially reduce their exposure to ad content.” Within the online advertising context, Cho and Cheon (2004) further suggested a three-component view of advertising avoidance, namely, cognition, affect and behavioral. They termed cognitive component of advertising avoidance as one’s belief about an object, and affective component of advertising avoidance as one’s feeling or reaction toward an object. This understanding of advertising avoidance in social media was also later accommodated by Kelly (2010) in understanding advertising avoidance in social media. Although acknowledging that one’s belief and one’s affect could contribute to the overall understanding of advertising avoidance as a construct, the current study argues that the cognitive and affective should be the antecedents of individuals’ ad avoidance behavior, which is supported by earlier studies’ supporting evidence that negative cognitive or affective evaluation would impact on consumers’ avoidance behavior (e.g., Alwitt & Prabhaker 1994). Moreover, even if considering the cognitive component in advertising avoidance, extant scholars held different perception upon the definition of the cognitive dimension of advertising avoidance. For example, Kelly et al. (2010) considered a common form of cognitive advertising avoidance as banner blindness (Hervet et al. 2011) which was referred as consumers’ intentional avoidance of looking at advertising banners online. While Cho and Cheon (2004) considered the cognitive component as one’s belief of an object. Comparing these two definitions of advertising 32 avoidance, one was more focusing on individuals’ viewing behavior (Kelly et al. 2010), while the other leans more toward the cognitive thinking (Cho and Cheon 2004). Therefore, given the current confusion in the definitions, it would be more sense-making if advertising avoidance is merely considered as a behavioral construct, while the cognitive and affective evaluation of advertising being considered as the cause of such behavioral outcomes. Following the behavior-focused understanding of advertising avoidance, another interesting thing to consider is the variation of advertising avoidance behaviors. In other words, advertising avoidance behaviors can be presented in various forms due to its embedded media context. For example, with TV commercials, consumers could avoid advertisement through ignoring the ad (cognitive avoidance), walking away (physical avoidance) or switching channels (mechanical avoidance) (Clancey 1994). While with online banner ads, consumers can also check off or install ad blockers in their browser to avoid receiving advertisements. In the context of social media, due to the wide range of social media platforms and their advertising product offerings, individuals’ advertising avoidance behaviors are quite diverse as well. Consumers could be intentionally ignoring, skipping, hiding advertisement, or installing ad blocker, etc. Thus, it would be difficult to measure all the advertising avoidance behaviors without a holistic definition of it. Comparing with the concept of consumer acceptance of social media advertising, consumers’ avoidance behaviors are mostly “active” behavior initiated by individuals. For example, when an individual intends to avoid advertisement in social media platforms, he/she would have to take further actions, such as “hide,” “check off,” “unfollow” or even pay a service fee to avoid seeing any types of sponsored advertisements. The amount of cognitive effort and cost is also much higher than simply receiving the advertisement. It is also assumed that one’s negative cognitive and affective evaluation of social media advertising is much higher as compared to when 33 receiving the advertisement. Therefore, in the current study, we consider advertising avoidance in social media as “consumers’ behavioral intention to withdraw from receiving any paid advertisement in the social media.” Enacted Affordances of Social Media and Consumer Responses to Advertising The fundamental function of the concept enacted affordances of social media is to reveal the specific characteristics/attributes of a social media vehicle in an individuals’ perception, which could further be used to understand how individuals respond to different messages on the specific social media vehicle. Such a theoretical logic falls in to the literature of media vehicle effect, also known as the vehicle source effect (Aaker and Brown 1972; Blair 1966; Woodside and Soni, 1990), which suggest that specific media vehicle, such as specific magazines, television programs, radio channels, could possess its own unique characteristics as perceived by the receiver, thereby induce different level of receptivity of information (Meenaghan & Shipley 1999). Therefore, different from the big picture of media effect, which would consider social media as a totally different type of media, apart from television, radio, press, the media vehicle effect considers the qualitative aspects of specific media vehicles in a more focused lens. Therefore, the enacted affordances of social media could be used to understand the specific media vehicle effect across different social media platforms. Specifically, the current study intends to investigate how such media vehicle effect would induce different levels of consumer responses to advertising feature in social media, using the enacted affordances of social media as an approach to analyzing the mechanism. To draw the relationship between the enacted affordances of social media and consumers’ responses to the advertising feature in social media, the researcher adopts a series of sociopsychology theories in the realm of information processing literature to elaborate the rationales. Because the enacted affordances of social media have a taxonomy that was built upon 34 two dimensions, namely, 1) the nature of the content and 2) the action on the content, the following section will be divided into two parts in proposing the hypotheses. The first part adopts the contextual priming effects in understanding how the nature of content in enacted affordances of social media could have an impact on consumers’ acceptance and avoidance of advertising in social media. While the second part adopts a dual-process mode in information processing models and limited cognition capacity theory to understand how the action on the content in enacted affordances of social media could lead to different consumer response to advertising. Yet, before laying out all the theoretical argument, the researcher intends to first further elaborate the nature of enacted affordances of social media as acquire knowledge of the individual that was formed through various types of learning experiences. Enacted Affordances of Social Media as Acquired Knowledge According to experiential learning theory (Kolb 1984, p.41), learning is “a process whereby knowledge is created through the transformation of experience. Knowledge results from the combination of grasping and transforming experience”. It is believed that learning is a holistic process of adaptation to the world that integrates thinking, feeling, perceiving and behaving. Through concrete experiences of learning, abstract conceptualization will be achieved and certain functional orientations will emerge for different objects/situations (Kolb 2014). In other words, individuals would form their own associated knowledge network for different objects/situations as a result of the adaptation process to the world. In the context of social media, such an experiential learning process also occurs. When an individual first gets in touch with a social media platform, he/she first need to build the associated knowledge network relates to the specific platform. Thus, intuitively, he/she will explore the affordances of the platform, such as understanding the main features of the 35 platform and also learn through observing norm regarding how others are using the platform. Through concrete use and reinforcement experiences of the social media platforms, individuals would gradually form the perception regarding how each social media platform could be used differently, and how each could satisfy their different gratifications. And based on these experiences, individuals would then build their own abstract conceptualization/knowledge about the affordances these social media platforms provide. Moreover, each individual would have their own learning outcomes of the social media affordances, since the formation of an individuals’ evaluation of the social media affordances, as mentioned earlier, are influenced by two major factors, i.e., the individual characteristics and the media characteristics. Therefore, variation in individuals’ differences and media characteristics could influence individuals’ actual user experience on the social media platform, which could further impact on the formation of abstract conceptualization of the social media platform’s affordances. For example, for certain individuals, the social media platform Instagram has high affordances of relational content consumption, while for other individuals who use Instagram mainly for artistic inspiration, they would consider Instagram as having high affordances of nonrelational content consumption. Naturally, the salience level of each type of enacted affordance in different social media platforms could also vary, since many of the enacted affordances types coexist. Taking the Instagram again as an example, if an individual follows both artist accounts and their social network accounts, then both the affordances of consuming relational content and nonrelational content could be high, or it could vary in terms of the salient level based on how many artists accounts and social accounts are being followed, as well as how often these accounts post content in the users’ Instagram newsfeed. 36 Therefore, as the outcome of individuals’ experiential learning of different social media, enacted affordances of social media, defined as “the acted-out action possibilities for a user to conduct on a social media platform to achieve the goals in a particular context”, can serve as a concept that not only detecting if social media platforms are different from each other based on the types of affordances, but also to what extent such differences are for each type of perceived affordances. Through such an understanding of the variation in social media affordance, scholars could adopt the concept to examine how each enacted affordance of social media could impact on individuals’ acceptance toward various kinds of messages. Specifically, as the interest of the current study, the researcher intends to draw the connection in between individuals’ enacted affordances of social media and their responses (i.e., acceptance and avoidance) to social media advertising. Since the enacted affordances of social media are categorized based on two dimensions: 1) the nature of the content and 2) the action on the content, it is thus believed that individuals could form their own associated knowledge network regarding what kind of content is more expected in certain social media platforms, as well as the associated knowledge network regarding what kind of behavior is more natural to certain social media platforms. Therefore, in the following section, the researcher would propose two sets of propositions, one featuring the dimension relating to the content nature (i.e., relational vs. non-relational) in the enacted affordances of social media to understand how it influence individuals’ acceptance and avoidance of advertising in social media; while the other would discuss the actions (i.e., consuming, contributing and creating) on the content in the enacted affordances of social media to detect how such behavioral tendencies in different social media platforms could impact individuals’ acceptance and avoidance of advertising in social media. 37 Contextual Congruence and Consumer Response Medium context, defined as “any number of form or content qualities associated with specific advertising mediums that are of interest to the researchers” (Jeong & Kim 2010), has been one of the major topics studied in the media effect literature in advertising. Specifically, scholars have focused on the concept of “congruence” in between the medium context and advertising messages (e.g., Aaker and Brown 1972; Goldberg & Gorn 1987; Yi 1990). In general, the “congruence” could be considered as a matching of advertisements and media vehicle platforms in terms of the featured cognitive attributes, the emotional tone, or the pathos execution style, etc. For example, in the contextual priming theory (Yi 1990a, 1990b), Yi proposed that if high congruence is achieved between the primed attributes in the context and the featured product attributes in the advertisement, the advertisement could result in a more positive brand evaluation. The “congruence” understood in the contextual priming focused more on the detailed contextual cues and the specific message attributes, which could be of great importance when investigating the message effects. However, in the current study, since the current focus leans more upon the media effects, the researcher intends to use a more macro lens to perceive advertising as a type of message, and discuss the general matching in between social media vehicles and advertising, through considering the original contextual environment formed by the general types of content being communicated on the social media platform. In other words, through considering the enacted affordances of social media from the “nature of content” dimension, the researcher intends to investigate what kind of social media vehicle would provide a better contextual environment for advertising messages as a whole. According to the schema theory, a schema refers to one’s “preexisting assumption about the way the world is organized.” The scheme theory, thus, discusses the information processing 38 process when new information is introduced (Axelrod 1973). It is believed that individuals’ have their own schemas of information to make sense of the complex environment in the world. To some extent, individuals’ formation of schema could also be considered as the result of experiential learning, since the learning process itself provides necessary information to form the associated knowledge network of the schema. When new information is introduced to one’s developed schema, individuals will evaluate the new information to see if it fits into the original pattern of information. Therefore, when a new type of information, such as an advertisement, is introduced into the social media platform, individuals would evaluate the new information to see if it fits into the context of the medium. As the growth of diversification of various social media platforms, each social media platform has their unique functions that users perceived as most relatable or useful. And due to the network structure differences, the type of content in social media for each individual could differ in many aspects. Such variation in social media’s functions, network structures, as well as content types could thus result in different individuals’ schema of the expected medium contextual environment. Enacted affordances of social media, serving as a construct to understand individuals’ perception of various social media platforms, thus, can be used to explain the impact of different expectation of contextual environment for each medium on individuals’ acceptance of new information, e.g., advertisement in social media. As mentioned earlier, one of the dimensions in enacted affordances of social media considers the nature of the content being either relational or non-relational. The relational content in social media is understood as content that is person-focused, made for initiating, maintaining or building the personal relationship on social media. While the non-relational content refers to content that is information-focused, which could be of either utilitarian or hedonic benefits. 39 Following the contextual congruence literature, as discussed earlier, the researcher argues that when the context of the social media platform is congruent with the newly introduced information type, i.e., advertisements, a more positive evaluation will be given toward ads, thus result in higher acceptance of ads. While, on the other hand, if the context of the social media platform is incongruent with the introduced stimuli, i.e., advertisements, a less positive evaluation will be given toward the ads, thus result in higher avoidance of the ads. Since social media platforms could be different in the level of enacted affordances relating to “relational content” and “non-relational content”, and paid advertisement is a part of non-relational content, the researcher, thus, suggests that social media platform with high affordances relating to non-relational content would be considered as more contextual-congruent for advertising messages, as compared to social media platforms with high perceived affordances relating to relational content. And the enacted affordances relating to non-relational content should have a greater impact on consumers on consumers’ acceptance of advertising messages in social media. Therefore, on the impact of the enacted affordances on consumer acceptance of social media advertising, the researcher hypothesizes the following: H1: The enacted affordances of non-relational content consumption, contribution and creation have a stronger impact on consumers’ acceptance of social media advertising, as compared to the enacted affordances relating to the relational content. Moreover, Kelly (2010) suggested that the lack of relevance of the advertising message in social networking sites could contribute to one’s avoidance of advertisement. And since social media platforms with high affordances in relational content are suggested as less contextualcongruent for advertising messages, the researcher proposes that consumers’ avoidance of the 40 advertising messages should be higher as compared to platforms that are contextual-congruent with non-relational content. And the enacted affordances relating to relational content should have a greater impact on consumers on consumers’ avoidance of advertising messages in social media. Therefore, on the impact of the enacted affordances on consumer avoidance of social media advertising, the researcher hypothesizes the following: H2: The enacted affordances of relational content consumption, contribution, and creation has a stronger impact on consumers’ avoidance of social media advertising, as compared to the enacted affordances relating to the non-relational content. Cognitive Capacity in Dual-Process Information Processing and Consumer Response The other dimension of the enacted affordances of social media considers the types of action individuals engage with the content, namely, consumption, contribution, and creation. Consumption of content mainly refers to individuals’ acquisition of various content on social media, while contribution refers to individuals’ response to others’ initiated content (e.g., liking, commenting), and creation refers to one’s own user-initiated content on social media. Extant theories in media studies have focused mostly on the information processing process, i.e., one’s information consumption/intake, but little theories were developed relating to one’s information output, such as information contribution and creation. Therefore, given the condition of lacking a direct theoretical mechanism that can distinguish the differences among consumption, contribution and creation, in the following section, the researcher would discuss the concepts of cognitive involvement and goal impediment to draw the connections in between the enacted affordances of social media and consumers’ responses to advertising in social media. 41 According to limited capacity model, individuals are information processors and their ability to process information are limited (Lang 2000). It assumes that the basic part of information process is to make sense of the stimuli and turn them into mental representations. Therefore, the entire process involves a group of information components which are being processed simultaneously. However, individuals’ mental resources are limited. Therefore, for each of the information component, the mental resources could be allocated differently. A closely related concept in information processing is the dual-process model, such as Elaboration Likelihood Model (Petty & Cacioppo 1986) and Heuristic-systematic model (Chaiken 1980, 1987). Both of these two dual-process models suggest that individuals process information through two routes/ in two ways, the central(systematically) route and peripheral (heuristically) route. When in the central route/systematic information processing, individuals engage in effortful information processing in forming their social judgments and making decisions. While in the peripheral route/heuristic information processing, individuals usually engage with effortless information processing. Individuals in this kind of information processing route are more relaxing as compared to the ones in the systematic information processing route. Yet, there is also a possibility of co-occurrence of both systematic (central) and heuristic(peripheral) information processing. Connecting the dual-process models with the enacted affordances of consuming, contributing and creating content in social media, there is a limitation because the foundation of the dual-process model is information processing (consuming content). However, the researcher argues that content contribution and creation are information processing behaviors that are switched form a passive mode to an active mode. When an individual has to contribute (e.g., like, comment) and create (e.g., post content), a series of information processing, such as processing information stimulus, information retrieval from structured knowledge network, and information 42 evaluation has to be taken place before one actually output any information, i.e., contributing and creating content in social media. Therefore, the process an individual has to go through before any information output is mostly systematic information processing, because it requires a fair amount of effort to make decisions of what and how an information output would be. While, consuming, as discussed earlier, takes two ways of information processing (e.g., central/systematic and peripheral/heuristic) because consuming content itself is the information processing process. Therefore, the actions of creating and contributing content majorly undergo systematic information processing, while the action of consuming content could take both kinds of information processing routes. And because systematic information processing demands more cognitive effort than heuristic information processing, the level of cognitive capacity thus would be higher when an individual is under heuristic information processing, while lower when an individual is under systematic information processing. Therefore, when advertising messages are introduced under one’s systematic information processing mode, the limitation of capacity could exclude any irrelevant information other than the central information of his/her interest. However, if the advertising message is presented under forced exposure, such a kind of message would intrude one’s ongoing information processing behavior, which could further evoke a higher level of perceived goal impediment of the advertisement. Scholars had long identified the perceived goal impediment as one of the most significant influencers of ad avoidance (Li et al. 2002; Cho & Cheon 2004; Predergast et al. 2014, Kelly 2010) when the individuals were conducting goal-oriented behavior on the Internet (Li et al. 2002). This is because goal-oriented behavior often requires a higher level of cognitive involvement or even a cognitive closure from other external stimuli. Therefore, when the advertisement intrudes the flow 43 of highly focused cognitive information processing, the negative affect would occur due to the goal impediment. Relating back to the action dimension of enacted affordances of social media, it is difficult to distinguish the amount of cognitive capacity that would be available among consuming, contributing, and creating content, because content consuming entails both the heuristic and systematic information processing. In other words, it is difficult to assure that the cognitive capacity differences between systematic consuming and creating in social media. However, the researcher argues that when an individual is undergoing heuristic information processing (content consuming), the amount of cognitive capacity is higher than when an individual is contributing or creating since both the contributing and creating content has to engage with a series of systematic information processing. And thus, would lead to a lower level of perceived goal impediment intrigued by the paid advertisements in social media. Therefore, the researcher proposes that: H3: The enacted affordance of content consumption (heuristically) has a stronger impact on consumer acceptance of social media advertising, as compared to content contribution and creation. H4: The enacted affordance of content creation has a stronger impact on consumer avoidance of social media advertising, compared to content consumption (heuristically) and content contribution. 44 CHAPTER 5 METHOD To sufficiently examine the proposed hypotheses in the current study, the researcher adopted a non-experimental, correlational approach using a single online survey instrument. The online survey was designed to examine the relationship between individual users’ enacted affordances of the social media platforms and their general acceptance and avoidance of advertisements on social media platforms. Such a non-experimental and correlational approach is believed to most readily (Vogt et al. 2014, Orcher 2014) for analyses of the relationships proposed between variables in the current study, i.e., identifying correlations between key variables, and comparing the level of impact between different variables on the same dependent variables. Study Design The online survey was designed beginning with a consent form which elaborates on the topic, the potential risk of participation, the protection of participants’ confidentiality and the compensation for participating. For the quality of the current data, participants who consented were first directed to a quality check question which asked about their willingness to commit to thoughtfully provide his/her best answers to each question in the survey. Only participants who committed to provide their best answers proceeded to the full survey. Moreover, the current survey also controlled for age and their actual use of social media, only individuals aged from 18 to 49 who are active social media users could proceed to full survey. The main survey starts with questions regarding individuals’ use of social media platforms. Each participant was asked to select four social media platforms they were actively using from eight options, namely, LinkedIn, Pinterest, Youtube, Yelp, Snapchat, Instagram, Twitter, and 45 Facebook. For each selected social media platform, a block of questions tailored to the specific platform was exposed to the participant. In each block of questions, participants answered questions relating to their individual usage of the platform, their evaluation of enacted affordances of social media, their information processing style when using the social media platform, their past experiences with advertising on the platform, as well as their general acceptance and avoidance of advertisements on the social media platform. Several relevant variables were also included in the survey, such as participants’ general evaluation of the ad value and relevancy of the ad on the platform, as well as their expectancy to see an ad on the platform. As an alternative measure of individuals’ advertising avoidance tendency, the survey also asked individuals’ “willingness to pay” in not having any advertisements on the social media platforms. The survey ended with questions regarding demographic information, including gender, ethnic background, education level, and annual household income in the year of 2016. Measurement Enacted affordances of social media were measured using the earlier developed scale in Chapter 3. The six types of enacted affordances were each measured using three statements. For example, “ [social media platform] is best for catching up with my social connections’ social life” for relational content consumption; “I always ‘like’ or ‘comment’ my social connections life-event on [social media platform]” for non-relational content consumption; “I use [social media platform] the most to share about my life events ” for relational content contribution; “I always see content that I find useful or entertaining on [social media platform]” for non-relational content consumption; “I always ‘like’ or ‘comment’ content I find useful or entertaining on [social media platform]” for relational content contribution and “ [social media platform] is where I share useful 46 or entertaining content the most” for non-relational creation. Participants were asked to evaluate the items for enacted affordances of social media through identifying their level of agreement with each statement using a 7-point Likert scale (1=strongly disagree, 7=strongly agree). Moreover, to distinguish the different information processing mode when individuals consume content on the social media platforms, an additional question addressing their level of cognitive effort was included. Specifically, we asked the participants to identify the percentage of time that he/she would be effortless processing the content and the percentage of time he/she would be highly involved with the content. Individuals’ acceptance of advertising was adopted from Wu’s (2016) scale in measuring one’s acceptance of advertising in the mobile social media application. Wordings of the items were revised to match the context of the current study. A total of five items were included in the measurement. Similarly, participants evaluated each statement on a 7-point Likert scale (1=strongly disagree, 7=strongly agree). Because the current study only interests in one’s general willingness to receive advertising in social media regardless of the actual advertising message, such measurement is of high appropriateness. Measurement for individuals’ avoidance of advertising was developed in Baek & Morimoto’s (2012) study in which they investigated individuals’ ad avoidance intention on personalized advertising. Wordings of the items were also revised to fit the current study’s scope. The final scale for advertising avoidance included five items/statements covered one’s intention to ignore advertisement on social media, attitude toward not having ad on social media, the behavioral intention in avoiding social media and their level of willingness in paying the platform in avoiding advertisement in social media. 47 Additionally, the researcher also added an alternative measurement of individuals’ advertising avoidance by asking their willingness to pay for not having paid advertisement for one month, which is measured by giving individuals a range of monetary value (i.e., 0 – 20 USD per month) he/she is willing to spend on advertising avoidance. Table 5 presents the measurement items in detail. Sample Description Participants of the current study were recruited through Qualtrics, a platform that allows participants to sign up studies in exchange for monetary reward. Each participant was given 5 dollars for their full participation. To ensure the quality of the findings, the age range of the extant study is set in between 18-49, which is in aligned with the fact that above 80% of this age group at least use one social media site. A quality check was added in the very beginning of the survey to ensure the participants’ commitment to provide genuine report of their answers. A total sample size of 574 participants was recruited in the current study, which satisfies the requirement of power analysis using G*Power3 to determine a sufficient size using an alpha of 0.05, a power of 0.80 and a small effect size (f=0.10) (Faul et al. 2009). All participants identified themselves as active social media platform users who met the screening requirement of the study participation. More than three fourth of the participants were female (N=451, 78.6%) and the rest identified themselves as male (N=123, 21.4%). Majority of the participants were Caucasian (N=431, 75.1%), followed by African American (N=57, 9.9%), Hispanic/Latino (N=47, 8.2%), Asian (N=24, 4.2%), Pacific Islander (N=3, 0.5%) and others (N=12, 2.1%). Because of the screening of age for the study participation, the participants’ age ranged from 18 to 49, with about 60% of the participants falls under the age of 34 (Millennials). Almost all participants had high school education (N=538, 99%) and 40% of the participants had a bachelor degree or higher. 48 Regarding the annual household income, more than half of the participants have had an annual household income greater than $70,000 (N=211, 36.9%) in the year of 2016, which is about the average household income as indicated in the latest U.S. Census Bureau data in 2014. Table 6 presents more specific information. Sample’s Social Media Selection Description Moreover, since each participant selected and evaluated four social media platforms in the survey, the final unit of analysis relating to the enacted affordances of social media resulted in a total of 2189 units (i.e., evaluations of social media platforms). Among all, majority of the participants selected Facebook (n=505, 23%), followed by Youtube (n=476, 22%), Pinterest (n=344, 16%), Instagram (n=286, 13%), Twitter (n=195, 9%), SnapChat (n=164, 7%), LinkedIn (n=117, 5%) and Yelp (n=102, 5%). The age median for these eight social media platforms are quite similar, falls into the range of 32 to 35, except SnapChat which is at a younger age demographic that has an age median at 27. Regarding the education level for the users across these eight social media platforms, the researcher found that people who selected LinkedIn has the highest percentage of receiving four-year college education among all (61.5%), followed by Yelp (48%), Twitter (46.7%), Facebook (40.8%), Instagram (40%), Pinterest (36.9%), Youtube (36.6%), and SnapChat (29.9%). Table 7 presents the details. 49 CHAPTER 6 DATA ANALYSIS AND HYPOTHESES TESTING The data analysis was conducted using the IBM SPSS version 24 and Amos. There are overall four main parts. The first part of the data analysis concerns factor validity and reliability, which included both exploratory factor analysis using SPSS and confirmatory factor analysis using Amos. The second part of the data analysis reported on the descriptive results regarding the differences of enacted affordances across various social media platforms. A hexagonal map that showcases the differences of enacted affordances for social media platforms was also generated. The third part focused on the statistical assumptions testing, which was made prior to the main hypotheses testing to ensure the legitimacy of the main data analysis. The main hypotheses testing first examined the existence of linear relationships between the enacted affordances of social media and individuals’ advertising acceptance and avoidance. And then the researcher made a comparison regarding the impact of various enacted affordances of social media on individuals’ acceptance and avoidance of social media advertising. Factor Validity and Reliability Test To prepare the data for further examination and analysis, the researcher first conducted factor analysis in testing the validity of the enacted affordances variables, followed by the reliability test using Cronbach’s α. The results showed that all the scales exceeded the recommended level of reliability at 0.7 (Hair et al. 1998), ranged from 0.912 to 0.972, which indicated a good reliability for the measurement. The factor loadings resulted from the factor analysis using maximum likelihood extraction with Promax rotation showed good factor loadings, which all satisfied the recommended level of 0.55 (Comrey and Lee 1992, Tabachnick and Fidell 2007). The confirmatory factor analysis also showed good factor loadings with a good model fit, 50 i.e., GFI=0.975, AGFI=0.965, NFI=0.991, CFI=0.994 and RMSEA=0.036 (Hu and Bentler 1999). See Table 8 for details. Enacted Affordances Differences of Social Media Platforms Adopting the measurement of six types of enacted affordances of social media, the descriptive results of the eight social media platforms in terms of their variation in enacted affordances are shown in Table 9. Moreover, Figure 1 also presents the illustration regarding the variation of the enacted affordances across the platforms, based on the current sample. As presented in Table 9 and Figure 1, all the social media platforms varied in their enacted affordances in the current sample. Specifically, Facebook could be considered as the big brother among all social media platforms for it is relatively more balanced among all enacted affordances, and it occupied the highest value for all the enacted affordances of social media (Consume_R=6.07, Contribute_R=5.44, Create_R=5.31, Consume_NR=5.56, Contribute_NR=5.33, Create_NR=5.27). Another two similarly balanced platforms in the current study are Instagram and Twitter. Instagram holds a relatively lower value across all enacted affordances (Consume_R=5.22, Contribute_R=5.05, Create_R=4.66, Consume_NR=5.16, Contribute_NR=5.01, Create_NR=4.46) as compared to Facebook, yet was evaluated as having more affordances than Twitter across all types of enacted affordances (Consume_R=4.34, Contribute_R=4.02, Create_R=3.74, Consume_NR=4.64, Contribute_NR=4.23, Create_NR=4.04). Although sharing the similar pattern with Facebook, Instagram, and Twitter in terms of enacted affordances of relational content (Consume_R=5.1, Contribute_R=3.75, Create_R=4.38), SnapChat holds less enacted affordances for non-relational content (Consume_NR=4.74, Contribute_NR=3.77, Create_NR=4.14) as compared to its relational 51 content affordances. Also, SnapChat has less affordances of contributing content, as compared to consuming and creating. Besides the above discussed balanced social platforms, the rest tend to have more focus on non-relational content, specifically for consumption. For example, YouTube has a high affordance of non-relational content consumption (Consume_NR=5.55) as compared to the other affordances (Contribute_NR=3.93, Create_NR=3.18, Consume_R=3.34, Contribute_R=3.15, Create_R=2.63). Two other social media platforms also share this pattern, namely, Pinterest and Yelp. Pinterest has a high affordance of non-relational content consumption (Consume_NR=5.5), but falls short on the others (Contribute_NR=3.8, Create_NR=3.75, Consume_R=2.94, Contribute_R=2.75, Create_R=2.37). Yelp also has a high affordance of non-relational content consumption (Consume_NR=4.49), and sits low on the other affordances (Contribute_NR=2.72, Create_NR=2.68, Consume_R=2.43, Contribute_R=2.33, Create_R=2.32). The last but not the least, the social media platform that holds the least affordances among all in the current sample is LinkedIn, which holds the lowest value across all in terms of the enacted affordances (Consume_R=2.87, Contribute_R=2.73, Create_R=2.25, Consume_NR=3.27, Contribute_NR=2.86, Create_NR=2.45). Advertising Acceptance and Avoidance Differences Across Social Media Platforms To examine the differences in advertising acceptance and avoidance among the eight social media platforms, the researcher conducted ANOVA analysis. Results indicated significant differences for advertising acceptance (F=3.998, p<.05) but not for advertising avoidance (F=0.921, p=0.489). Table 10 presents the descriptive results. Figure 2 shows the pattern for advertising acceptance. 52 As indicated in the descriptive result in Table 10, Facebook has the highest advertising acceptance (M=3.75, SD=1.92) among all social media platforms, followed by Instagram (M=3.52, SD=1.90), Twitter (M=3.46, SD=2.01), Youtube (M=3.45, SD=1.86), Yelp (M=3.31, SD=2.00), Pinterest (M=3.21, SD=1.83) and LinkedIn (M=3.07, SD=1.80). Regarding individuals’ advertising avoidance, there were no significant differences across all social media platforms. The mean values of advertising avoidance for all social media platforms ranging from 4.62 (Yelp) to 4.99 (Youtube). Because the ANOVA results successfully laid a solid foundation for the existence of advertising acceptance differences across various social media platforms, it makes the hypotheses testing more relevant. Therefore, to further understand the impact of enacted affordances of social media on individuals’ acceptance of social media advertising, as well as finding potential impactful enacted affordances on individuals’ avoidance of social media advertising, the following part would continue with the hypotheses testing and data exploration. Statistical Analysis Assumption Examination Before testing hypotheses, the researcher first examined the data to ensure it satisfies all the assumptions of running a linear regression using the current data. To satisfy the required assumption for linear regression, the variables should meet the following rules: 1) linear relationship, 2) normal distribution, 3) no or little multicollinearity, 4) no auto-correlation, 5) homoscedasticity. The linear relationships and homoscedasticity were examined through investigating scatter plot in SPSS, the results indicated proper scatter plot graph for the linear relationship and the homoscedasticity of the data. The normal distribution was examined through investigating the skewness and kurtosis of the current data. The results showed that the skewness of independent variables and dependent variables fell into the range of -0.880 to 0.231, the kurtosis 53 of the variables was in the range of -1.517 to 0.007, which satisfied the acceptable range of -2 to 2 according to George and Mallery (2010). Regarding autocorrelation, since the residuals are independent of each other, there is no autocorrelation in the current data sample. To test the multicollinearity issue, the researcher first ran correlations among the independent variables (i.e., enacted affordances of social media) with the dependent variable (i.e., advertising acceptance and avoidance). Results are presented in Table 11. According to the results presented in Table 11, some of the independent variables that were supposed to be compared in its impact have high correlations, such as the enacted affordance of consuming relational content is highly correlated with the enacted affordance of contributing relational content (β=0.819), the enacted affordance of consuming relational content is highly correlated with the enacted affordance of creating relational content (β=0.825), and the enacted affordance of contributing relational content is also highly correlated with the enacted affordance of creating relational content (β=0.830). Such a high correlation among the enacted affordances of relational content was also spotted when the scale was first developed. Although the wordings of the items were linguistically distinct from each other, and there is a legit reason why these enacted affordances of relational content are highly correlated, to satisfy the statistical assumption, the researcher further calculated the variance inflation factor (VIF) to detect the issue of multicollinearity. According to the Belsey & Welsch (2005), VIF value higher than 10 would be considered as having multicollinearity issue. After running collinearity diagnosis, the results indicated that all the VIF value for three enacted affordances of relational content were under 10. Specifically, the VIF value for relational content consumption, contribution and creation were separately at 3.826, 3.930, and 4.036. Therefore, although the enacted affordances of relational content were highly correlated, there was no multicollinearity issue among these three variables. 54 Hypotheses Testing Because the proposed hypotheses focus on comparing the impact of individual predictors on the dependent variables, a statistical method that can compare the beta coefficients of individual predictors in multiple regression is needed. Therefore, the researcher adopted the method developed by Cumming (2009) which uses 95% confidence interval to interpret the significance of the differences between two beta coefficients. Specifically, Cumming (2009) demonstrated that when the corresponding 95% confidence intervals of the two predictors overlap by not more than 50%, then the two beta coefficients are likely to be statistically significantly different from each other. Figure 3 from Cumming (2009) help explain the method visually. As indicated in Figure 3, when the 95% confidence interval of the two estimates have an overlap less than 59%, then it is considered the impact from these two individual predictors are significantly different at the level of p=0.05; and when the 95% confidence interval of the two estimates have an overlap less than 14%, we could consider the beta coefficients of the two individual predictors are significantly different at the level of p=0.01; while when there is no overlap in between these two estimates’ 95% confidence interval, then we could say that the impact from these two individual predictors are significantly different at the level of p=0.001. By and large, the 50% overlap is the acceptable cutoff for a statistically significant difference between two beta weights. The procedure of the analysis started with standardizing all the independent variables and dependent variables into Z-scores and ran multiple regressions using the standardized value of independent and dependent variables with bootstrapping. Since the proposed hypotheses address the influence differences between enacted affordances of non-relational content and relational content, as well as among enacted affordances of content consumption, contribution, and creation 55 on the dependent variables (i.e., advertising acceptance and avoidance), the researcher adopted two methods in comparing the differences. The following section presents the two means in comparing the impact. Analysis Approach 1: Enacted Affordances as Individual Measures Hypothesis 1 proposed that the enacted affordances relating to non-relational content would have a stronger impact on consumers’ acceptance of social media advertising as compared to the enacted affordances relating to the relational content. When considering each enacted affordance as individual measures, this hypothesis can be translated into 3 pairs of comparison, namely, 1) the enacted affordance of relational content consumption vs. non-relational content consumption, 2) the enacted affordance of relational content contribution vs. non-relational content contribution and 3) the enacted affordance of relational content creation vs. non-relational content creation. Multiple regression analyses were conducted to detect the standardized beta weights and its level of significance. Results showed that all six types of enacted affordances had a significant relationship with advertising acceptance. The standardized beta weights for the enacted affordance of relational content consumption is at 0.319, and the non-relational content consumption is at 0.219. The standardized beta weights for the enacted affordance of relational content contribution is at 0.252, and the non-relational content consumption is at 0.274. Moreover, the standardized beta weights for the enacted affordance of relational content creation is at 0.255, and the non-relational content creation is at 0.266. Table 12 presents more details for the upper and lower value of 95% confidence interval. To test the hypothesis that the impact of enacted affordance of relational content consumption (β=0.319) and the non-relational content consumption (β=0.219) on consumer 56 acceptance of advertising is statistically significantly different from each other, their corresponding 95% confidence intervals were estimated via bias-corrected bootstrap (1,000 resamples). As mentioned before, when the confidence intervals of the two estimates overlapped by less than 50%, the beta weights would be considered as statistically different from each other (p<.05; Cuming, 2009). As shown in Figure 4a, the two types of enacted affordance have a certain area of overlap, yet direct conclusion regarding if the overlapping area is less than 50% cannot be drawn. Therefore, to evaluate the hypothesis precisely, half of the average of the overlapping confidence intervals were calculated (0.0385) and added to the enacted affordance of relational content consumption beta weight lower bound estimate (0.278), which yielded 0.3165. As the enacted affordance of non-relational consumption upper bound estimate of 0.255 was smaller than the value of 0.3165, the difference between the enacted affordance of the relational content consumption and the non-relational content consumption standardized beta weight (Δβ = 0.1) was considered significantly different from each other. To compare the difference between the enacted affordance of relational content contribution and non-relational content contribution, the same procedure was conducted as the first pair of comparison (i.e., enacted affordance of relational content consumption vs. non-relational content consumption). The researcher calculated the half of the average of the overlapping confidence intervals and resulted in the value of 0.0625, adding it to the enacted affordance of nonrelational content contribution beta weight lower bound estimate (0.213) yielded 0.2755. And since the enacted affordance of relational contribution upper bound estimate of 0.316 exceeded the value of 0.2755, the difference between the enacted affordance of relational content contribution and non-relational content contribution standardized beta weight (Δβ = 0.22) was not significant. 57 The same procedure was again conducted to compare the enacted affordance of relational content creation and non-relational content creation. The half of the average of the overlapping confidence intervals and resulted in the value of 0.067. After adding it to the enacted affordance of non-relational content creation beta weight lower bound estimate (0.197), it yielded the value of 0.264. The enacted affordance of relational creation upper bound estimate of 0.320 exceeded the value of 0.264, thus, the difference between the enacted affordance of relational content creation and non-relational content creation standardized beta weight ( Δβ = 0.11) was not significant. Overall, the results showed that the enacted affordance of relational content consumption was more impactful than the enacted affordance of non-relational content consumption on influencing individuals’ advertising acceptance. And there was no significant difference between the enacted affordances of relational content contribution and non-relational content contribution in influencing one’s acceptance of social media advertising. Nor there was a significant difference in between the enacted affordance of relational content creation and non-relational content creation. Therefore, based on the above findings, H1 was not supported. Hypothesis 2 proposed that the enacted affordances of relational content consumption, contribution and creation would have a stronger impact on consumers’ avoidance of social media advertising as compared to the enacted affordances relating to non-relational content. Similarly, when considering each enacted affordance as individual measures, the translated version of this hypothesis was again the 3 pairs of comparison: 1) the enacted affordance of relational content consumption vs. non-relational content consumption, 2) the enacted affordance of relational content contribution vs. non-relational content contribution and 3) the enacted affordance of relational content creation vs. non-relational content creation. 58 However, before comparing the beta coefficient, the results of multiple regression with Zscores showed that only the enacted affordances of non-relational content consumption had a significant relationship with the advertising avoidance (β=0.172, p<.05). The other enacted affordances of social media didn’t show a significant relationship with individuals’ advertising avoidance. Table 13 presents the details. Therefore, the fundamental base was not met for the beta coefficient comparison. Thus, H2 was not supported. Hypothesis 3 proposed that for social media platforms that are mostly under heuristic information processing, the enacted affordances of content consumption would have a stronger impact on consumers’ acceptance of social media advertising, as compared to contribution and creation. Therefore, before detecting the existence of significant relationship between the enacted affordances of social media and individuals’ acceptance of advertising, the data were separated based on the level of heuristic processing of information (i.e., High vs. Low, cutoff point=4). To test the proposed hypothesis 3 and 4, the researcher would only use the data in high heuristic information processing to compare the enacted affordance of content consumption with content contribution and creation. When considering each enacted affordance as individual measures, the translated version of this hypothesis would be comparing 1) the enacted affordance of relational content heuristic consumption with contribution and creation; and 2) the enacted affordance of non-relational content heuristic consumption with contribution and creation. Multiple regression analyses were conducted to investigate if the enacted affordances had relationship with individuals’ acceptance of advertising. In addition, to compare the impact differences, the values of 95% confidence interval were also retrieved. As indicated in Table 14, the enacted affordance of relational content consumption has an insignificant relationship (β = 0.054, p = 0.125) with individuals’ advertising acceptance in social media, when regressed with 59 the enacted affordance of relational content contribution and creation (β = 0.011, p = 0.761). Therefore, the comparison was only made among the pairs related to the enacted affordances of non-relational content consumption, contribution, and creation. Figure 5a shows the beta coefficient comparison of the enacted affordance of non-relational content consumption and contribution using 95% confidence intervals. Since there is no overlap in between the confidence intervals of the enacted affordances of non-relational content consumption and contribution, we can conclude easily that there is a significant difference between the two beta coefficients, as the lower bound of non-relational content contribution is at 0.366, while the upper bound of non-relational content consumption is at 0.144, which made the distance between the lower bound of non-relational content contribution and the non-relational content consumption at 0.222. Therefore, the difference between the two beta coefficients was significant. In other words, the enacted affordance of non-relational content contribution was statistically more influential (Δβ = 0.37) than the enacted affordance of non-relational content consumption in affecting individuals’ advertising acceptance in social media. Similarly, to make the comparison regarding the impact on individuals’ advertising acceptance in between the enacted affordance of non-relational content consumption and creation, the same method is adopted. As indicated in figure 5b, there is obviously no overlap in between the confidence intervals between the enacted affordance of non-relational content consumption and creation. The lower bound of non-relational content creation is at 0.354, while the upper bound of non-relational content consumption is at 0.180, which led to the difference between these two confidence intervals at 0.174. Therefore, the findings revealed that there is a significant difference of impact between the enacted affordance of non-relational creation and consumption (Δβ = 60 0.336), and the enacted affordance of non-relational content creation was more influential than the enacted affordance of non-relational consumption on individuals’ acceptance of advertising. In comparing the impact difference between the enacted affordance of non-relational content contribution and creation, the same procedure described above was repeated. As indicated in the figure 5c, the two confidence intervals are almost identical, which showed the possibility that there is no significant difference in between the two kinds of enacted affordances. Yet, to provide a more scientific comparison, the half of the average of the overlapping confidence intervals was calculated which was at 0.0715. After adding it to the lower bound value of enacted affordance of non-relational content contribution, the value was 0.2945. And since the upper bound value of the enacted affordance of content creation was 0.314, which was larger than the value of 0.2945. Therefore, the difference of the impact in between the enacted affordance of non-relational content contribution and creation was not significant. In general, since the current comparison was not able to involve the enacted affordances of relational content, and the enacted affordance related to non-relational content (i.e., consumption, contribution, and creation) were found as having the opposite direction in terms of their influential impact, i.e., the enacted affordance of non-relational content consumption was found as less influential than the enacted affordance of non-relational content contribution and creation. Thus, the hypothesis 3 was not supported. Hypothesis 4 proposed that the enacted affordance of content creation would have a stronger impact on consumer avoidance of social media advertising, as compared to heuristic content consumption and content contribution. Once again, when considering the enacted affordances of social media as individual measures, the analyses included comparing 1) the enacted affordance of relational content creation with heuristic consumption and contribution; and 61 2) the enacted affordance of non-relational content creation with heuristic consumption and contribution. Multiple regression analyses were, again, conducted to detect the foundation of comparing coefficient beta values --- the existence of a significant relationship between the enacted affordances of social media and individuals’ advertising avoidance. Table 15 presents the details regarding the regression results, and it also includes the confidence interval values. Results in Table 15 showed a rather mixed result with several comparisons not able to proceed due to the insignificant relationship in between some of the enacted affordances and individuals’ advertising avoidance. For example, when regressed with the enacted affordance of relational content consumption, the enacted affordance of relational content creation has no significant relationship with one’s advertising avoidance. Moreover, in comparing the enacted affordance of relational content contribution and relational content creation, both two enacted affordances were not significant. The same results were also shown in the comparison of nonrelational content contribution and non-relational content creation. Therefore, due to these insignificant results, the comparison process was not able to proceed. Thus, H4 was not supported. The current method of analysis considered the enacted affordances of social media as individual measure as developed in the earlier scale development study. Although all the hypotheses were not supported, the result yielded interesting findings. However, the limitation of this method is the cumbersome process in finishing all the comparison, as well the requirement for the existence of a significant relationship among all the enacted affordances of social media with the dependent variables. Therefore, the researcher proposed a second method in the data analysis, which is to adopt Geometry method in calculating the areas for relational and non-relational affordances, as well as the areas for content consumption and content creation affordances. The following section will elaborate the method in detail. 62 Analysis Approach 2: Enacted Affordances as Calculated Area The computation method of using area as representation of variables was rarely seen in the field of media studies. However, in the field of endocrinology or neurosciences, scholars have often employed such method (Pruessner et al. 2003). It is believed that such a method of computing area of related variables allows simplifying the statistical analysis procedure without sacrificing information from the individual measures. For the current study, the area method not only can help simplifying the steps as occurred in the section of method 1 but also it could avoid the situation when single measure hinders the overall analysis. Therefore, the researcher adopts such a method of understanding 1) the impact of relational vs. non-relational content affordances, as well as the 2) the impact of content consumption vs. contribution vs. creation affordances in social media on advertising acceptance and avoidance, a geometry formula is first needed to compute the area of the relational content, non-relational content, content consumption and content creation. Figure 4 serves as an example for the explanation of the computation. Since the enacted affordances form a regular hexagon shape, and there is a total of six triangles in the hexagon shape, as well as six inside angles. Thus, each of the inside angles would be 360° 2 = 60°. Considering the measures of enacted affordances of social media as the edges of the triangles, we thus have the edges value as well as the inside angle value. To calculate the total area of a triangle, the formula is presented as follows. A= 1 a. b ∗ Sin60° 2 A represents the area of the triangle, a, b each represents the edges by the inner angle, which was calculated as 60°. And adapting this formula to the computation of area for relational content, 63 the researcher first calculated the triangle area formed by the edges of the enacted affordance of relational content consumption and relational content contribution, and then the area for the triangle formed by the edges of enacted affordance of relational content contribution and relational content creation. To calculate the total area of relational content, the researcher added the area of these two triangles. Therefore, :;<=>?@AB>=CAB??< ∗ KLM60° 2 To calculating the area for non-relational content, the same approach was adopted. The researcher first calculated the area for the triangle by the edges of non-relational content consumption affordance and non-relational content contribution affordance, and adding it with the triangle area formed by the edges of non-relational content contribution affordance and nonrelational content creation affordance, to get the total area for non-relational content affordance. :OAB;<=>?@AB>=CAB??< ∗ KLM60° 2 When considering the enacted affordances relating to consumption, contribution and creation of content, the researcher used the enacted affordances of relational content consumption and non-relational content consumption as the two edges of a triangle to calculate the area of content consumption, used the enacted affordances of relational content contribution and nonrelational content contribution as the two edges of a triangle to calculate the area of content contribution and used enacted affordances of non-relational content creation and relational content 64 creation as the two edges to calculate the area for content creation, which resulted in the following two formulas. :CAB??@AB = 1 D ∗ PDEI<>?< ∗ KLM60° 2 EI<>?< After the computation of these five areas, namely, enacted affordance for relational content, enacted affordance for non-relational content, enacted affordance for content consumption, enacted affordance for content contribution and enacted affordance of content creation, the researcher tested the hypotheses using these five area variables. In testing the hypotheses using these area variables, the method of Cumming (2009) was adopted again in comparing the impact from these independent predictors on the dependent variables. Therefore, multiple regression analyses with transformed Zscores of the area variables were conducted with the values of 95% confidence interval retrieved. The results of the standardized coefficient beta weights and the confidence intervals for testing H1 and H3 are all presented in Table 16 (DV=AdAcceptance), while the results of the standardized coefficient beta weights and the confidence intervals for testing H2 and H4 are all presented in Table 17 (DV=AdAvoidance). To test the Hypothesis 1 which examines the impact difference between the enacted affordance of relational content and non-relational content on individuals’ acceptance of advertising, the half of the average of the overlapping confidence intervals was calculated, which resulted in the value of 0.066. After adding it to the lower bound estimate of enacted affordance 65 of non-relational content 0.241, the result yielded the value of 0.307, which was larger than the upper bound estimate of enacted affordance of relational content 0.296. Thus, the difference between the two beta coefficients was considered significant. In other words, the enacted affordance of non-relational content was statistically more influential (Δβ = 0.083) than the enacted affordance of non-relational content in affecting individuals’ advertising acceptance in social media. Figure 7a illustrates this result in a high-low chart. Thus, H1 was supported. To test hypothesis 2 which examines the impact difference between the enacted affordance of relational content and non-relational content on individuals’ avoidance of advertising, the researcher adopted the same approach. However, due to the insignificant correlation in between the enacted affordance of relational content to individuals’ avoidance of social media advertising, as shown in Table 16, the comparison was not able to proceeded. Yet, the enacted affordance of non-relational content was found having a significant impact on one’s avoidance of social media advertising (β = 0.080, p = 0.037). In general, H2 was not supported. To test hypothesis 3, the data was first split into high heuristic information processing and low heuristic information processing. Only the data in the high heuristic information processing was included in the final analysis (N=1759). In comparing the impact of enacted affordances of content consumption, content contribution, and content creation on consumers’ acceptance toward social media advertising, several multiple regressions were conducted first to detect if significant relationships exist in between these variables with consumers’ advertising acceptance, shown in Table 16. Results presented that 1) when comparing the enacted affordances of content consumption with content contribution, both two kinds of enacted affordances were shown significantly influencing consumers’ acceptance (β = 0.116, p = 0.002; β = 0.382, p = 0.000); 2) when comparing the enacted affordances of content consumption with content creation, only 66 the enacted affordance of content creation has a significant relationship with consumers’ acceptance of social media advertising (β = 0.439, p = 0.000), but not for the enacted affordance of content consumption (β = 0.062, p = 0.107); 3) when comparing the enacted affordances of content contribution with content creation, both types of enacted affordances also showed significant relationships with consumers’ acceptance towards social media advertising ( β = 0.216, p = 0.000; β = 0.306, p = 0.000) . Therefore, the comparison in between the beta coefficients can only be made in between the enacted affordances of content consumption and contribution, and enacted affordances of content creation and contribution. In comparing the enacted affordance of content consumption and content contribution, the scholar drew the high-low chart for the 95% confidence intervals for the two variables. As shown in figure 7b, there is no overlap in between the confidence intervals of the enacted affordances of content consumption and content contribution. The lower bound of the enacted affordance of content contribution is at 0.306, while the upper bound of enacted affordance of content consumption is at 0.200, which has a distance of 0.106. Therefore, the beta coefficients of the enacted affordance of content consumption and the enacted affordance of content contribution are significantly different from each other. In other words, the enacted affordance of content contribution is more influential than the enacted affordance of content consumption regarding consumers’ advertising acceptance in social media (Δβ = 0.266). Similarly, the high-low chart is again created for the comparison of the enacted affordance of content contribution and content creation. As indicated in figure 7c, there is a certain level of overlap in-between these two variables’ confidence intervals. Therefore, the precise calculation was made to identify if there is over 50% overlap in between the two variables’ confidence intervals. Half of the average of the overlapping confidence intervals was calculated (0.0485) and 67 added to the enacted affordance of content contribution lower bound estimate (0.216), which yielded 0.2645. As the enacted affordance of content creation upper bound estimate 0.320 exceeded the value of 0.2645, the difference between the enacted affordance of content contribution and enacted affordance of content creation on consumers’ advertising acceptance is not significant. In conclusion, H3 was only partially supported because the enacted affordance of content contribution is significantly more influential than the enacted affordance of content consumption, but no comparative analysis can proceeded with the enacted affordance of content creation and content consumption. However, the enacted affordance of content creation, when comparing with the enacted affordance of content consumption is at a beta value of 0.439, which is quite high in terms of an influential predictor. Thus, the researcher suggests that when a social media platform is majorly under the heuristic information processing mode, the H3 is partially supported. To test hypothesis 4 which addressed the differences in between heuristic consumption affordance and creation affordance in individuals’ avoidance of social media advertising, the issue of insignificant correlation in between enacted affordance of social media and the dependent variable was again spotted, as presented in Table 17. Therefore, the comparison was not able to proceed. H4 was not supported. In conclusion, adopting this area computation method, hypothesis 1 was supported, H3 was partially supported, while hypothesis 2, and 4 were not supported. Yet, the researcher is also curious if when not just considering the heuristic information processing condition, would there be any difference. Therefore, the following analysis was further conducted to examine if all social media platforms, regardless being used majorly for heuristic information processing or not, would yield interesting findings. Therefore, the regression analyses between enacted affordances of 68 content consumption, contribution creation with one’s advertising acceptance were conducted. Results showed all relationships to be significant, which can be proceeded for further comparison, as indicated in Table 18. Using the same method of comparing their confidence intervals’ overlapping area, the confidence interval comparison results further showed that there is no overlapping between the confidence intervals of enacted affordance of content contribution and content consumption, as well as the enacted affordances of content creation and content creation, as indicated in figure 8a and 8b. Thus, the researcher can conclude that there is a significant difference between the coefficients of enacted affordances of content contribution and content consumption ( Δβ = 0.220). In other words, the enacted affordance of content contribution is more influential than the enacted affordance of content consumption in addressing consumers’ acceptance of social media advertising. Moreover, there is also a significant difference between the coefficients of enacted affordances of content creation and content consumption ( Δβ = 0.256), i.e., the enacted affordance of content creation is also more influential than content consumption in affecting consumers’ acceptance of social media advertising. However, there is no significant difference regarding the influences from the enacted affordances of content contribution and creation, because there is an over 50% overlap in between their confidence intervals, as shown in figure 8c. 69 CHAPTER 7 DISCUSSION & CONCLUSION The current study adopted two approaches in examining the proposed hypotheses, one approach considered the measurement of enacted affordances of social media as single items, while the other approach uses the calculated area score for each kind of enacted affordance. Generally speaking, the hypotheses relating to advertising avoidance (H2 and H4) were not able to be examined due to the non-existence of linear relationships among the various kinds of enacted affordances of social media and individuals’ advertising avoidance. Therefore, only the hypotheses relating to advertising acceptance (H1 and H3) were able to be examined. Results in analysis approach 1 showed insignificant opposite findings as the researcher predicts. When comparing the impact from the enacted affordances of relational content and nonrelational content using the instrument as individual measures, the results revealed that there was no statistical difference between the enacted affordance of relational content contribution and nonrelational content contribution, and the enacted affordance of relational content creation and nonrelational content creation. Although there was a significant difference in between the beta coefficients of the enacted affordance of relational content consumption and non-relational content consumption, the enacted affordance of relational content was found as more impactful in influencing one’s social media advertising acceptance than the enacted affordance of nonrelational content (Δβ=0.1,p<0.5). While using the analysis approach 2 (i.e., calculated area score), the enacted affordance of non-relational content was found slightly more influential than the enacted affordance of relational content on individuals’ advertising acceptance, which was in align with the researchers’ prediction. The area for relational content and the area for non-relational content, results showed that the 70 enacted affordance of non-relational content was more impactful than the enacted affordance of relational content (Δβ=0.083,p<0.5). Interpreting these findings from a holistic point of view, the enacted affordance of nonrelational content of social media is more influential than the enacted affordance of relational content. However, the difference between the difference in between these two enacted affordances of social media is quite small, which make the insignificant difference between the enacted affordance of relational and non-relational content contribution, as well as the enacted affordance of relational and non-relational content creation understandable. Moreover, the enacted affordance of relational content consumption was actually being found as more influential than the enacted affordance of non-relational content consumption on one’s acceptance of advertising in social media. Therefore, although the second method statistically proved the enacted affordance of nonrelational content was more impactful than the enacted affordance of relational content on one’s acceptance of social media advertising, after the overall evaluation, the researcher suggests that the impact from the enacted affordance of relational content and non-relational content do not differ much. But the enacted affordance of non-relational content does have a slightly stronger impact on consumers’ acceptance of social media advertising as compared to the enacted affordance of relational content. The effect of contextual congruence, based on the current study’s categorization of relational and non-relational content, seems to have its limitation in its predictive power on individuals’ responses to advertising. One possibility could be that the salience of the relational content vs. non-relational content difference is not particularly emphasized in the current study design, since we are using an online survey as an instrument which only asks for participants general evaluation rather than having them directly experience the content. Moreover, based on 71 the results of this sample, many of the social media platforms resulted in a quite balanced shape of relational content and non-relational content. In other words, many of the social media platforms had no extreme enacted affordances. Therefore, such a comparison between the enacted affordance of relational content and non-relational content were not quite distinct given current samples’ use of social media, which could, as well, lead to the slight difference result. The researcher suggests that if such differences could be made as more salient across different social media platforms, the results could be more obvious. Secondly, the researcher also compared the impact of the enacted affordance of content consumption, contribution, and creation on individuals’ advertising acceptance. Both results from two methods suggested that the enacted affordance of content creation had a higher impact on individuals’ advertising acceptance than the enacted affordance of content consumption. Method 1 showed that both enacted affordances of non-relational content contribution and creation had a stronger influence than one’s advertising acceptance (Δβ = 0.37; Δβ = 0.336) than the nonrelational content consumption. And the enacted affordances associated with relational content was not able to be compared due to the lack of significant relationships. While in Method 2, the general comparison resulted in a higher influence from the enacted affordance of content creation as compared to the enacted affordance of content consumption Δβ = 0.256 . Such a finding contradicted the researcher’s original hypothesis which proposed that the enacted affordance of content consumption (heuristic) would be more influential than the enacted affordance of content creation in advertising acceptance because individuals would have more cognitive capacity to allocate on heuristic information (e.g., advertisements, peripheral cues). However, such a hypothesis was built upon the logic that considers the actual activity states. In other words, when developing the hypothesis, the researcher was considering the enacted 72 affordance of social media as the actual activity that one is undergoing. However, given the current study’s design, the survey was asking about one’s general perception and evaluation of the social media use as well as his/her attitude and behavioral intention toward accepting or avoiding advertisement in social media. Therefore, what the current findings indicate is that the enacted affordances of content contribution and content creation would have a more influential impact than the enacted affordance of content consumption in accepting the social media advertising. In other words, when a social media platform has an increase in its affordances of content contribution/creation, its users would be willing to accept the social media advertising at a larger range than when increasing a social media platform’s enacted affordance of content consumption. According to Schivinski et al.’s (2016) study about consumer engagement measurement which included the three activity factors – consumption, contribution, and creation, they found that these three dimensions from the lower to higher levels correspond to one’s engagement with the social media brand-related content. Therefore, the researcher suggests that the enacted affordances of content contribution and content creation could also be considered as corresponding to the media engagement level. In other words, the value of enacted affordance of content contribution and creation indicates the value of media engagement. The higher the enacted affordances of content contribution and creation, the higher the media engagement. Calder et al. (2009) set the stage that connected the media engagement with online advertising effectiveness and found that one’s media engagement experience is positively associated with the advertising effectiveness. Specifically, they identified two kinds of engagement --- Personal engagement and Social-interactive engagement. And their quasi-experiment finding indicated that the Social-interactive engagement had a significant impact on one’s reaction to advertisement even after controlling for the personal engagement. In other words, one’s interactive 73 user experiences with a medium platform could influence one’s reaction to advertisements. Therefore, since the enacted affordance of content contribution and creation indicates higher level of media engagement, and one’s engagement with an online medium could positively influence one’s reaction to advertisement, the current finding that the enacted affordance of content contribution and creation were found having a stronger impact one’s acceptance of advertising in social media is of no surprise. Additionally, prior media studies have also proved such correlation between media vehicle engagement with individuals’ positive responses to advertising (e.g., Coutler 1998; DePelsmacker et al. 2002; Nicovich, 2005; Wang 2006). The last but not least, another interesting finding in the current study is the insignificant relationships between the enacted affordances of social media and one’s advertising avoidance. Presumably, we considered that the advertising acceptance and advertising avoidance are two sides of one coin. However, based on the findings of the current study, such a pre-assumption seems problematic. In the current study, the researcher considers advertising avoidance in social media as “consumers’ behavioral intention to withdraw from receiving any paid advertisement in the social media.” And it was measured using items such as “I intentionally ignore any paid advertisement” and “I would ‘hide’ or ‘check off’ paid advertisement”. Such behavioral intention seems to be more of an “active” engagement with the advertisement than a “passive” ignorance of the social media advertisement. The researcher, thus, suggests that there could be a differentiation in between active avoidance and passive avoidance, such as the banner blindness found in earlier online banner advertising studies (Drèze and Hussherr 2003; Chatterjee 2008). Moreover, prior scholars’ exploration of regarding the antecedents of consumers’ advertising avoidance included perceived goal impediment, perceived ad clutter, and prior negative experiences (Kelly et al. 2010; Cho and Cheon 2004). Therefore, the research suggests that future studies could consider using 74 the enacted affordances of social media as a way to categorize social media platforms, which further lead the social media type as a moderator building on the extant model of advertising avoidance. Although almost none of the hypotheses in the current study was supported as proposed, the results yielded interesting and proving findings that help provides insightful results for understanding the relationship between the enacted affordances of social media and individuals advertising acceptance of social media advertising. Moreover, the results also directed many potential future studies that could help further untie the usefulness of enacted affordances of social media as a concept in the social media literature. The last chapter will discuss the contributions, implications, limitations of the current research project with more detailed description of future directions. 75 CHAPTER 8 CONTRIBUTIONS, LIMITATIONS, AND FUTURE DIRECTION The contribution of the current research project has three-fold, namely, theoretical contribution, methodological contribution, and managerial contribution. The theoretical contribution and methodological contribution can be further implied for academic implications. And the managerial contribution could be further used for industry implications. In this chapter, the researcher would discuss each type of contribution and its implications. The researcher would also discuss the limitations and provide directions for future studies. First of all, to the researcher’s knowledge, the current research project takes a pioneering approach to understand the differences across various social media platforms in their enacted affordances. Its major theoretical contribution relies on the proposition of the concept “enacted affordance” which incorporating both the media characteristics and individual characteristics in understanding the medium factors in the communication process. Moreover, the taxonomy that categorizes six types of enacted affordance of social media provides a new theoretical lens for communication scholars to compare social media platforms regarding the types of enacted affordances and its power in driving communication outcomes. For example, scholars adopting this concept and its measurement could examine a specific kind of content, such as health-related content in health communication, new-related content in journalism, sports-related content in sports communication, so that scholars could find out which social media platform has the highest enacted affordance of a specific kind of content. Communication scholars could also understand how individuals interact with the content on different social media platforms, and how these interactions could lead to different communication outcomes. 76 Secondly, the current research also innovates the academic practices regarding the method section, specifically, the data analysis method. Traditional communication scholars, when operationalizing construct, often adopts Likert scale, which are items that measured using 7 points or 5 points. However, in the current study, the researcher introduces the method of computed area as a mean to get an area value for the specific construct by using single-item measurement. Such a method has been adopted in the endocrinology field (Pruessner et al. 2003), but it was rarely seen in the communication realm. Therefore, the current study provides an empirical example of the application of the area computation method in the field of communication studies. Thirdly, the current study also provides various managerial contributions that could be further implied in the industry. An intuitive one is that marketing/advertising practitioners in social media field could consider adopting the taxonomy of the enacted affordances of social media in tailoring and developing their social media communication strategies. It is believed that for each type of enacted affordances of social media, practitioners could use it as a point for divergent thinking. For example, for social media platforms perceived by individuals as having high affordances of relational content creation, practitioners could consider how to develop communication strategy or campaigns that help these individuals creating relational content, while engaging with brands or organizations that the practitioner represents for. In other words, after knowing what are the enacted affordances of social media platforms for various individuals, practitioners could consider how to add value to or be compatible with one’s use of the social media platforms. As for managerial implications regarding the various social media technology companies, there are also several implications for the company’s sustainable growth regarding its advertising revenue sustainability. For instance, the findings indicted that the enacted affordance of non- 77 relational content was found having a stronger impact than the enacted affordance of relational content in driving individuals’ acceptance of social media advertising. Therefore, to increase users’ acceptance or willingness to receive advertisements on a social media platform or balancing the use of original social media content and advertising content, the social media platform could consider increasing the amount of non-relational content being interacted upon on the platform. Additional, an alternative approach could also be to expand services to non-relational-contentfocused social media platform in driving more effective advertising responses and revenue. Moreover, the results also revealed that social media platforms that have high media engagement, i.e., high enacted affordances of content contribution and content creation from individuals’ evaluation would be more likely to have high consumer acceptance of advertisements on the social media, as compared to the influence from enacted affordance of content consumption. Thus, the researcher suggests that, in increasing the effective advertising response, social media platforms should encourage users to be more actively engaged with the content on the platform through the design of the user interaction/experience, such as provide more intervention in encouraging users to initiate, share or comment on relational and non-relational content. Limitations and Future Studies The current research project also has several limitations that should be noted. First of all, the literature review section only reviewed articles that were related to the field of “social media advertising” but not in public relations, computer-mediated communication, and other media studies. The future study could consider expanding the current review scope to a wider range of literature to see if there are other medium-focused studies that are related to the various affordances of social media platforms. Moreover, future researches could be conducted to investigate the antecedents of the formation of the proposed six types of enacted affordances of social media. 78 Secondly, although controlled the quality of the sample’s response through Qualtrics, the final sample set was primarily female participants. The gender differences could be a reason that majority of the social media platforms were quite balanced in terms of both relational and nonrelational content because women were found are more engaged with social media use than men (PewInternet, 2017). Therefore, future studies could consider collecting more data from male to test the hypotheses again and see if there would be any differences. Thirdly, the current study design adopted a quite general way in measuring the involved variables without introducing actual messages to make a more salient effect. Therefore, in the future studies, the researcher could adopt experimental design as an alternative method to investigate the proposed hypotheses and examine if increasing the salience of certain variables would make a difference in the result. Or more external factors could be introduced to the comparative studies of social media platforms, such as message factors, individual differences, and source influences, to detect if such a hybrid medium-factor is more of a mediator or a moderator in achieving the communication effectiveness. Finally, the researcher would like to use the metaphor of “playing table pool” to make the closing remarks of the current study. The empirical study of the current project is like the ‘first pole” of the game. Since studies regarding social media medium comparison are at its early stage, many of the questions and issues are unknown to the scholars. Just as found in the current study, most of the proposed hypotheses were not supported, yet an opposite result was found significant. Therefore, through the theoretical lens of enacted affordances of social media, the researcher believes that both scholars and practitioners could find their own angles in getting their answers of social media communication. 79 APPENDIX 80 Table 1. Statistics of Published Social Media Advertising Research from 2006 – 2016 Journal Name Journal of Interactive Advertising Journal of Advertising Research Journal of Advertising Journal of Current Issues and Research in Advertising International Journal of Advertising Journal of Marketing Journal of Consumer Research Journal of Marketing Research Journal of Interactive Marketing Journal of Marketing Communication Total Total (Empirical Studies Only) Total Relevant Articles 27 28 4 3 21 10 0 11 35 20 159 133 81 Medium 5 3 0 0 1 1 0 0 2 1 13 10% Individual 21 13 3 2 16 1 0 5 20 8 89 67% Source 4 0 1 2 6 1 0 2 13 5 34 26% Message 8 6 1 3 6 3 0 2 11 5 45 34% Table 2. Thematic Categorization of Social Media Advertising Studies from 2006 - 2016 Input Factors Theme Sub-categories Studies Individual Factors Motivation 1) motivation to use social media 2) motivation to engage with brand 3) motivation to engage with advertisement Gangadharbatal (2008), Elsenbeiss et al. (2011);Phillips et al. (2014); Sashittal and Jassawalla (2015); Muk & Chung (2014); Kwon et al. (2014); Taylor et al. (2012); Tsai and Men (2013); Hsieh and Chang (2015); Hayes et al. (2014); Hayes et al. (2016); Individual Characteristics 1) Influence of demographic and psychological differences Hoy & Milne (2010); Do et al. (2014); Kim and Yoon (2014); Goodrich & Mooij (2014); Yeo (2012); Mai & Olsen (2015); Chu et al. (2015); Kamal et al. (2013); Fang et al. (2016); Consumer Brand Relationship 1) consumers’ self-identification with brand 2) consumers’ interaction with brand Yeh (2011); Do et al. (2014); Shan and King (2015); Hamilton et al. (2016); Labrecque (2013); Wang et al. (2011); Shan and King, (2015); Colliander et al. (2012); Labrecque (2013); Hamilton et al. (2016). Moon et al. (2013); Concern, Avoidance, & Attitude 1) privacy concern in social media 2) Consumers' skepticism toward advertising message 3) the link between attitude and behavioral outcomes Jeong and Coyle (2014); Kelly et al. (2010); Rozendaal et al. (2013); van Noort et al. (2014); Jung et al. (2015); Tucker (2014); Ham (2016); Huang et al. (2012); Alhabash et al. (2015); Prendergast et al. (2010); General Source Type Comparison 1) User generated content 2) Firm generated content or marketer generated content 3) Celebrity generated content Steyn et al. (2011); Hautz et al., (2013); Morris et al. (2016); Wood and Burkhalter (2014); Kumar et al. (2016) Social Information of Source 1) relationship with a source (e.g., tie-strength) 2) perception of the source (e.g., perceived expertise) Pan and Chiou 2011; Chu and Kim 2011; Want et al. 2011; van Noort et al. 2012; Shan & King 2015; Hayes et al. 2016; Dubois et al. (2016); Thompson & Malaviya 2013, Paek et al., 2015; Walker 2012;Jin & Phua 2014; Phua & Joo 2016; Source Factors 82 Table 2. (cont’d) Message Factors Influencer Identification Identification and categorization of influencers Morrison et al. (2013); Araujo et al. (2016); Chatterjee (2011); Thompkins (2011); Katona (2011); Jung et al. 2015; Message Valence & Its Interaction Combination with Message Factors Pan & Choiu (2011); Corstjens & Umblijs’ (2012); Daughtery & Hoffman’s (2014); Jin & Phua (2014) Message Analysis 1) Factors driving consumer engagement 2) Factors driving purchase intention, memory and attitude De Vries et al. 2012; Chen & Lee 2014; Kim & Yoon 2014; Rooderkerk & Paulwels (2015); Theo et al. (2015); Reichelt et al. 2014; Keyzer et al. 2015; Edlira Shehu et al. 2016; Van-tien Dao et al. (2014); Kononova & Yuan (2015);Luna-Nevarez & Torres (2015); Malte et al. (2015); Vander Bergh et al. (2011); Kumar et al. (2016); 3) Categorization of message characteristics Content Analysis Medium Factors 1) the typologies of social media advertising 2) the typologies of a specific kind of advertisement in social media 3) social media advertisement on different types of social media platform Stephen & Galak (2012); Jung et al. (2015); Bang & Lee (2016); Grant et al. (2015); Lueck (2015); Vargo (2016); Smith et al. (2012); Nelson-field et al. (2013); Phillips et al. (2014); Christian et al. (2014) Comparison of Social Media and Traditional Media Trusov et al. (2009); Pfeiffer & Zinnbauer (2010); Mabry & Porter (2010); Spotts et al. (2014); Pynta et al. (2014) Comparison of Different Social Media Platforms Smith et al. (2012); Van-Tien Dao et al. (2014);Logan (2014); Individual Relationship with Medium Lee & Ahn (2013); VanMeter et al. (2015); Wu (2016) A Macro-Approach to Social ‘Medium’ Seraj (2012); Peters et al. (2013); 83 Table 3. Typologies of Enacted Affordances of Social Media Relational Non-Relational Consume The extent to which an individual has taken actions of reading relational content on the social media. Contribute The extent to which an individual has taken actions of responding to relational content on the social media. Create The extent to which an individual has taken actions of sharing and posting relational content on the social media. The extent to which an The extent to which an The extent to which an individual has taken individual has taken individual has taken actions of reading actions of responding actions of sharing and non-relational content to non-relational posting non-relational on the social media. content on the social content on the social media. media. 84 Table 4. Finalized Scale Items for Enacted Affordances of Social Media Consume Relational Content ______ is best for catching up with my social connections' social life. ______ is best for learning about my social connections’' life. ______ is where I can get updates of my friends' life activities the most. Contribute Relational I always "like" or "comment" my social connections' life-event updates on ______. Content I always "like" or "comment" my social connections’ post about their life on ______. I always "like" or "comment" my social connections' posts about themselves on ______. Create Relational I use ______ the most to share about my life events. Content ______ is the channel on which I post the most of my life activities. I post about my life on ______ regularly. Consume Non-relational I always see content that I find useful or entertaining on ______. Content I expect to see entertaining or useful media content on ______ regularly. I often see useful or entertaining information on ______. Contribute NonI always "like" or "comment" content I find useful or entertaining on ______. relational Content If encountering useful or entertaining content on ______, I always "like" or "comment" it. It is common for me to "Like" or "Comment" useful or entertaining content on ______. Create Non-relational ______ is where I share useful and entertaining content the most. Content I always post media content that I find useful or entertaining on ______. If encountering useful or entertaining content, I always share it on ______. 85 Table 5. Measurement Table Construct Items Consume Relational Content [Social Media Name] is best for catching up with my social connections' social life. Yang (2016) [Social Media Name] is best for learning about my social connections’' life. [Social Media Name] is where I can get updates of my friends' life activities the most. I always "like" or "comment" my social connections' life-event updates on [Social Media Name]. I always "like" or "comment" my social connections’ post about their life on [Social Media Name]. I always "like" or "comment" my social connections' posts about themselves on [Social Media Name]. I use [Social Media Name] the most to share about my life events. [Social Media Name] is the channel on which I post the most about my life activities. I post about my life on [Social Media Name] regularly. I always see content that I find useful or entertaining on [Social Media Name]. I expect to see entertaining or useful media content on [Social Media Name] regularly. I often see useful or entertaining information on [Social Media Name]. I always "like" or "comment" content I find useful or entertaining on [Social Media Name]. If encountering useful or entertaining content on [Social Media Name], I always "like" or "comment" it. It is common for me to "Like" or "Comment" useful or entertaining content on [Social Media Name]. [Social Media Name] is where I share useful and entertaining content the most. I always post media content that I find useful or entertaining on [Social Media Name]. If encountering useful or entertaining content, I always share it on [Social Media Name]. I am willing to receive advertising on [Social Media Name] in the future. Wu (2016) I would pay attention to advertising message I receive on [Social Media Name] in the future. My general intention to accept advertising in [Social Media Name] is very high. I will think about reading advertising in [Social Media Name]. I will accept advertisement on [Social Media Name] in the future. I intentionally ignore any paid advertisement on [Social Media Name] Baek & Morimoto (2012) I usually don't pay attention to paid advertisement on [Social Media Name] It would be better if there were no paid advertisement on [Social Media Name] I would "hide" or "check off" paid advertisement on [Social Media Name] I would be willing to pay [Social Media Name] to not have any paid advertisement. (deleted) New Please identify the percentage of time you are just effortless browsing content on [Social Media Name] Contribute Relational Content Create Relational Content Consume Non-relational Content Contribute Non-relational Content Create Non-relational Content Acceptance of Advertisement Avoidance of Advertisement Heuristic Information Processing Source 86 Table 6. Sample Description Measure Item Count Percentage Gender Male 123 21.40% Female 451 78.60% 18-24 113 19..69% 25-29 94 16.38% 30-34 128 22.30% 35-39 106 18.47% 40-44 63 10.96% 45-49 70 12.20% Middle School 6 1.0%% High School/GED 124 21.60% Some College 137 23.90% 2-year College Degree 76 13.20% 4-year College Degree 153 26.70% Master Degree 71 12.40% Doctoral Degree 7 1.2%% White/Caucasian 431 75.10% Asian 24 4.2%% Hispanic/Latino 47 8.2%% Black/African American 57 9.9%% Pacific Islander 3 0.50% Others 12 2.10% less than $10,000 64 11.10% $10,000 - $19,999 43 7.50% $20,000 - $29,999 71 12.40% $30,000 - $39,999 56 9.80% $40,000 - $49,999 37 6.40% $50,000 - $59,999 54 9.40% $60,000 - $69,999 37 6.40% $70,000 - $79,999 42 7.30% $80,000 - $89,999 25 4.40% $90,000 - $99,999 33 5.70% $100,000 - $149,999 67 11.70% More than $150,000 45 7.80% Age Education Ethnicity Household Income (2016) 87 Table 7. Selection/Use of Social Media Platforms Platform LinkedIn Pinterest Youtube Yelp SnapChat Instagram Twitter Facebook Number of evaluation 117 344 476 102 164 286 195 505 Percentage 5% 16% 22% 5% 7% 13% 9% 23% 88 Age Median 35 33 32 34 27 32 35 34 4-year College Education 61.50% 36.90% 36.60% 48% 29.9%% 40% 46.70% 40.80% Table 8. Factor Analysis and Reliability Test Constructs EFA loadings CFA loadings R_Consumption 0.954 0.935 0.905 0.711 0.681 0.652 0.682 0.63 0.57 0.917 0.896 0.817 0.930 0.901 0.884 0.94 0.841 0.664 0.947 0.943 0.952 0.959 0.964 0.955 0.910 0.953 0.959 0.904 0.892 0.848 0.943 0.954 0.943 0.918 0.914 0.913 R_Contribution R_Creation NR_Consumption NR_Contribution NR_Creation 89 Crochbach's Alpha 0.963 0.972 0.958 0.912 0.963 0.939 Table 9. Enacted Affordances of Social Media Platforms Platform N Linkedin 117 Pinterest 344 Youtube 476 Yelp 102 SnapChat 164 Instagram 286 Twitter 195 Enacted Affordances of Social Media Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR 90 Mean 2.87 2.73 2.25 3.27 2.86 2.45 2.94 2.75 2.37 5.5 3.8 3.75 3.34 3.15 2.63 5.55 3.93 3.18 2.43 2.33 2.32 4.49 2.72 2.68 5.1 3.75 4.38 4.74 3.77 4.14 5.22 5.05 4.66 5.16 5.01 4.46 4.34 4.02 3.74 4.64 4.23 4.04 Std. Deviation 1.84 1.97 1.87 1.74 1.98 1.84 1.85 1.89 1.81 1.53 2.05 1.93 2.17 2.11 2.05 1.45 2.11 2.09 1.97 1.92 1.96 2.06 2.06 2.08 1.74 2.03 1.92 1.8 2.04 1.93 1.62 1.72 1.86 1.62 1.75 1.88 1.92 2.04 2.07 1.78 2.06 2.02 Table 9 (cont’d) Facebook 505 Consume_R Contribute_R Create_R Consume_NR Contribute_NR Create_NR 91 6.07 5.44 5.31 5.56 5.33 5.27 1.28 1.64 1.7 1.4 1.67 1.64 Figure 1. Enacted Affordances for Each Social Media Platform 92 Table 10. ANOVA Analysis Result for Advertising Acceptance and Avoidance Construct Platform N Mean Ad_Acceptance LinkedIn Pinterest Youtube Yelp SnapChat Instagram Twitter Facebook 117 344 476 102 164 286 195 505 3.07 3.21 3.45 3.31 3.12 3.51 3.46 3.74 Std. Deviation 1.80 1.83 2.01 2.02 1.90 2.01 1.93 1.91 Ad_Avoidance LinkedIn Pinterest Youtube Yelp SnapChat Instagram Twitter Facebook 117 344 476 102 164 286 195 505 4.91 4.82 4.99 4.62 4.86 4.83 4.76 4.85 1.66 1.74 1.74 2.02 1.97 1.74 1.71 1.76 93 F Sig. 3.998 0.000 0.921 0.489 Figure 2. Advertising Acceptance Across Social Media Platforms 94 Figure 3. Beta Coefficient Comparison using 95% Confidential Intervals 95 Table 11. Correlation Analysis for Independent Variables Consume _R Contribute _R Pearson 1 .819** Correlation Pearson Contribute_R .819** 1 Correlation Pearson Create_R .825** .830** Correlation Pearson Consume_NR .394** .425** Correlation Pearson Contribute_NR .648** .793** Correlation Pearson Create_NR .721** .768** Correlation **. Correlation is significant at the 0.01 level (2-tailed). Consume_R 96 Create_ R Consume _NR Contribute _NR Create _NR .825** .394** .648** .721** .830** .425** .793** .768** 1 .383** .676** .812** .383** 1 .584** .514** .676** .584** 1 .770** .812** .514** .770** 1 Table 12. Standardized Coefficients and Confidence Interval Results_H1 t Sig. Zscore(Consume_R) Zscore(Consume_NR) Standarrdized Coefficients Beta 0.319 0.219 15.351 10.549 Zscore(Contribute_R) Zscore(Contribute_NR) 0.252 0.274 Comparison Zscore(Create_R) 0.255 Zscore(Create_NR) 0.266 a. Dependent Variable: Zscore(Adacceptance) 95% Confidence Interval Lower Upper 0.000 0.000 0.278 0.181 0.363 0.255 8.273 8.98 0.000 0.000 0.189 0.213 0.316 0.334 8.008 8.345 0.000 0.000 0.193 0.197 0.32 0.326 97 Figure 4a. Beta Coefficient Comparison of Enacted Affordances of Relational Content Consumption and Non-relational Content consumption using 95% Confidential Intervals 98 Figure 4b. Beta Coefficient Comparison of Enacted Affordances of Relational Content Contribution and Non-relational Content Contribution using 95% Confidential Intervals 99 Figure 4c. Beta Coefficient Comparison of Enacted Affordances of Relational Content Creation and Non-relational Content Creation using 95% Confidential Intervals 100 Table 13. Standardized Coefficients and Confidence Interval Results_H2 t Sig. Zscore(Consume_R) Zscore(Consume_NR) Standardized Coefficients Beta 0.003 0.172 0.114 7.495 0.909 0.000 Zscore(Contribute_R) Zscore(Contribute_NR) -0.007 0.049 -0.215 1.463 0.830 0.144 1.091 0.298 0.275 0.765 Zscore(Create_R) 0.040 Zscore(Create_NR) 0.011 a. Dependent Variable: Zscore(AdAvoidance) 101 Table 14. Standardized Coefficients and Confidence Interval Results_H3 t Sig. Zscore(Consume_R) Zscore(Contribute_R) Zscore(Consume_NR) Zscore(Contribute_NR) Standardized Coefficients Beta 0.054 0.424 0.067 0.415 1.534 11.989 2.638 16.383 0.125 0.000 0.008 0.000 -0.015 0.359 0.025 0.378 0.129 0.503 0.125 0.485 Zscore(Consume_R) Zscore(Create_R) Zscore(Consume_NR) Zscore(Create_NR) 0.011 0.466 0.086 0.422 0.304 12.621 3.56 17.441 0.761 0.000 0.000 0.000 -0.065 0.394 0.046 0.386 0.084 0.549 0.141 0.48 6.491 7.693 7.171 8.8 0.000 0.000 0.000 0.000 0.167 0.212 0.175 0.223 0.316 0.356 0.314 0.364 Comparison Zscore(Contribute_R) 0.237 Zscore(Create_R) 0.281 Zscore(Contribute_NR) 0.233 Zscore(Create_NR) 0.285 a. Dependent Variable: Zscore(AdAcceptance) 102 95% Confidence Interval Upper Lower Figure 5a. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Consumption and Contribution using 95% Confidential Intervals 103 Figure 5b. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Consumption and Creation using 95% Confidential Intervals 104 Figure 5c. Beta Coefficient Comparison of Enacted Affordances of Non-relational Content Contribution and Creation using 95% Confidential Intervals 105 Table 15. Standardized Coefficients and Confidence Interval Results_H4 t Sig. Zscore(Consume_R) Standarrdized Coefficients Beta 0.129 3.225 0.001 0.083 0.363 Zscore(Contribute_R) -0.105 -2.630 0.009 - 0.330 -0.040 Zscore(Consume_NR) Zscore(Contribute_NR) 0.159 - 0.067 5.646 -2.364 .000 .018 0.186 -0.222 0.418 -0.017 Zscore(Consume_R) Zscore(Create_R) Zscore(Consume_NR) Zscore(Create_NR) 0.093 -0.059 0.154 -0.064 2.216 -1.404 5.688 -2.363 0.027 0.166 0.000 0.018 0.010 -0.238 0.180 -0.199 0.308 0.048 0.391 -0.019 -1.183 1.3191 0.727 -0.245 0.237 0.165 0.468 0.806 -0.233 -0.049 -0.090 -0.151 0.075 0.237 0.185 0.107 Comparison Zscore(Contribute_R) -0.050 Zscore(Create_R) 0.058 Zscore(Contribute_NR) 0.027 Zscore(Create_NR) -0.009 a. Dependent Variable: Zscore(AdAvoidance) 106 95% Confidence Interval Upper Lower Figure 6. Example for the Computation of Enacted Affordance Areas 107 Table 16. Standardized Coefficients and Confidence Interval Results_Area_H_Accept Standardized Coefficients Beta 0.224 0.307 t Sig. 6.785 9.293 0.000 0.000 0.154 0.241 0.296 0.375 Zscore(H_ConsumptionArea) Zscore(H_ContributionArea) 0.116 0.382 3.173 10.402 0.002 0.000 0.042 0.306 0.200 0.462 Zscore(H_CreationArea) Zscore(H_ConsumptionArea) 0.439 0.062 11.479 1.613 0.000 0.000 0.353 -0.017 0.517 0.141 5.477 7.762 0.000 0.000 0.134 0.216 0.320 0.389 Comparison Zscore(RelationalArea) Zscore(Non-relationalArea) Zscore(H_ContributionArea) 0.216 Zscore(H_CreationArea) 0.306 a. Dependent Variable: Zscore(AdAcceptance) 108 95% Confidence Interval Upper Lower Table 17. Standardized Coefficients and Confidence Interval Results_Area_H_Avoid Standardized Coefficients Beta -0.001 0.080 t Sig. -0.038 2.088 0.970 0.037 -0.078 0.009 0.067 0.159 Zscore(H_ConsumptionArea) Zscore(H_ContributionArea) 0.135 -0.073 3.230 -1.759 0.001 0.079 0.044 -0.156 0.220 0.019 Zscore(H_CreationArea) Zscore(H_ConsumptionArea) -0.063 0.127 -1.435 2.906 0.151 0.04 -0.138 0.038 0.026 0.205 -0.010 0.972 0.331 0.992 -0.099 0.047 0.091 0.130 Comparison Zscore(RelationalArea) Zscore(Non-relationalArea) Zscore(H_ContributionArea) 0.000 Zscore(H_CreationArea) 0.044 a. Dependent Variable: Zscore(AdAvoidance) 109 95% Confidence Interval Upper Lower Figure 7a. Beta Coefficient Comparison of Enacted Affordances of Relational Content and Nonrelational Content using 95% Confidential Intervals 110 Figure 7b. Beta Coefficient Comparison of Enacted Affordances of Content Consumption and Content Contribution under Heuristic Condition using 95% Confidential Intervals 111 Figure 7c. Beta Coefficient Comparison of Enacted Affordances of Content Contribution and Content Creation under Heuristic Condition using 95% Confidential Intervals 112 Table 18. Standardized Coefficients and Confidence Interval Results_Area_All_Accept Comparison Standardized Coefficients Beta t Sig. 95% Confidence Interval Upper Lower Zscore(ContributionArea) Zscore(ConsumptionArea) 0.363 0.143 10.667 4.213 0.000 0.000 0.287 0.070 0.444 0.214 Zscore(CreationArea) Zscore(ConsumptionArea) 0.382 0.126 11.210 3.707 0.000 0.000 0.313 0.059 0.456 0.200 Zscore(ContributionArea) Zscore(CreationArea) 0.242 0.280 6.783 7.850 0.000 0.000 0.162 0.196 0.331 0.357 a. Dependent Variable: Zscore(AdAcceptance) 113 Figure 8a. Beta Coefficient Comparison of Enacted Affordances of Content Consumption and Content Contribution using 95% Confidential Intervals 114 Figure 8b. Beta Coefficient Comparison of Enacted Affordances of Content Creation and Content Consumption using 95% Confidential Intervals 115 Figure 8c. Beta Coefficient Comparison of Enacted Affordances of Content Contribution and Content Creation using 95% Confidential Intervals 116 Table 19. Standardized Coefficients and Confidence Interval Results_Area_All_Avoid Comparison Standardized Coefficients Beta t Sig. 95% Confidence Interval Upper Lower Zscore(ConsumptionArea) Zscore(ContributionArea) 0.152 -0.070 3.931 -1.796 0.000 0.073 0.67 -0.157 0.245 0.004 Zscore(CreationArea) Zscore(ConsumptionArea) -0.037 0.125 -0.940 3.204 0.347 0.001 -0.118 0.041 0.042 0.206 Zscore(ContributionArea) Zscore(CreationArea) -0.002 0.069 -0.038 1.680 0.969 0.093 -0.087 -0.015 0.087 0.152 a. Dependent Variable: Zscore(AdAcceptance) 117 REFERENCES 118 REFERENCES Aaker, D. A., & Brown, P. K. (1972). Evaluating vehicle source effects. Journal of Advertising Research, 12(4), 11-16. Alwitt, L. F., & Prabhaker, P. R. (1994). Identifying who dislikes television advertising: Not by demographics alone. Journal of Advertising Research, 34(6), 17-30. Anastasi, A. (1976). Psychological Testing. Macmillan Publishing Co. New York. Apter, M. J. (1989). Reversal theory: Motivation, emotion and personality. Taylor & Frances/Routledge. Axelrod, R. (1973). Schema theory: An information processing model of perception and cognition. American Political Science Review, 67(04), 1248-1266. Bærentsen, K. B., & Trettvik, J. (2002, October). An activity theory approach to affordance. In Proceedings of the second Nordic conference on Human-computer interaction (pp. 5160). ACM. Belsley, D. A., Kuh, E., & Welsch, R. E. (2005). Regression diagnostics: Identifying influential data and sources of collinearity (Vol. 571). John Wiley & Sons. Blair, W.S. (1966), ``Attitude research and the qualitative value of magazines'', in Adler, L. and Crespi, I. (Eds), Attitude Research at Sea, American Marketing Association, Chicago, IL. BIA/Kelsey on Statista (2017). Social media advertising spending in the United States from 2011 to 2018 (in billion U.S. dollars). Retrieved from https://www.statista.com/statistics/246341/social-media-advertising-spending-in-the-us/ Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on psychological science, 6(1), 3-5. Bullmore (1975) What is Advertising? Advertising Association UK. Retrived from http://www.adassoc.org.uk/publications/what-is-advertising/ Bullmore (2016) What is Advertising? Advertising Association UK. Retrived from http://www.adassoc.org.uk/publications/what-is-advertising/ Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230. Brown, T. (2006). CFA with equality constraints, multiple groups, and mean structures. Confirmatory factor analysis for applied research, 236-319. 119 Calder, B. J., Malthouse, E. C., & Schaedel, U. (2009). An experimental study of the relationship between online engagement and advertising effectiveness. Journal of Interactive Marketing, 23(4), 321-331. Casaló, L. V., Flavián, C., & Guinalíu, M. (2008). Promoting consumer's participation in virtual brand communities: A new paradigm in branding strategy. Journal of marketing Communications, 14(1), 19-36. Clancey, M. (1994). The television audience examined. Journal of advertising Research, 34(4), S1-S1. Chatterjee, P. (2008). Are unclicked ads wasted? Enduring effects of banner and pop-up ad exposures on brand memory and attitudes. Journal of electronic commerce Research, 9(1), 51. Chemero, A. (2003). An outline of a theory of affordances. Ecological Psychology, 15(2), 181– 195. Chen, G.M. (2011). Tweet this: A uses and gratifications perspective on how active Twitter use gratifies a need to connect with others. Computers in Human Behavior. 27(2), 755–762. Cho, C. H., & Cheon H. (2004). Why do people avoid advertising on the internet? Journal of advertising, 33(4), 89-97. Chu, S. C., & Kim, Y. (2011). Determinants of consumer engagement in electronic word-ofmouth (eWOM) in social networking sites. International journal of Advertising, 30(1), 4775. Cialdini, R. B. (1993). Influence: The psychology of persuasion. New York: Morrow. Comrey AL and Lee HB (1992) A first course in factor analysis (2nd edition). Hillsdale,NJ: Lawrence Erlbaum Associates. Cook, J. D. (1981). The experience of work: A compendium and review of 249 measures and their use. London; New York: Academic Press. Correa, T., Hinsley, A. W., & De Zuniga, H. G. (2010). Who interacts on the Web?: The intersection of users’ personality and social media use. Computers in Human Behavior, 26(2), 247-253. Corstjens, M., & Umblijs, A. (2012). The power of evil. Journal of Advertising Research, 52(4), 433-449. Coulter, K. S. (1998). The effects of affective responses to media context on advertising evaluations. Journal of Advertising, 27(4), 41-51. 120 Courtois, C., Mechant, P., De Marez, L., & Verleye, G. (2009). Gratifications and seeding behavior of online adolescents. Journal of Computer Mediated Communication, 15(1), 109–137. Cumming, G. (2009). Inference by eye: reading the overlap of independent confidence intervals. Statistics in medicine, 28(2), 205-220. Daugherty, T., & Hoffman, E. (2014). eWOM and the importance of capturing consumer attention within social media. Journal of Marketing Communications, 20(1-2), 82-102. Davis, F. D. 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quart, (13) 319–339. De Keyzer, F., Dens, N., & De Pelsmacker, P. (2015). Is this for me? How consumers respond to personalized advertising on social network sites. Journal of Interactive Advertising, 15(2), 124-134. De Pelsmacker, P., Geuens, M., & Anckaert, P. (2002). Media context and advertising effectiveness: The role of context appreciation and context/ad similarity. Journal of Advertising, 31(2), 49-61. De Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand fan pages: An investigation of the effects of social media marketing. Journal of Interactive Marketing, 26(2), 83-91. Do, H., Ko, E., & Woodside, A. G. (2015). Tiger Woods, Nike, and I are (not) best friends: how brand's sports sponsorship in social-media impacts brand consumer's congruity and relationship quality. International Journal of Advertising, 34(4), 658-677. Donath, J. (2007). Signals in social supernets. Journal of Computer-Mediated Communication, 13(1), 231–251. Dubois, D., Bonezzi, A., & De Angelis, M. (2016). Sharing with Friends Versus Strangers: How Interpersonal Closeness Influences Word-of-Mouth Valence. Journal of Marketing Research, 53(5), 712-727. Drèze, X., & Hussherr, F. X. (2003). Internet advertising: Is anybody watching? Journal of interactive marketing, 17(4), 8-23. Dwyer, P. (2007). Measuring the value of electronic word of mouth and its impact in consumer communities. Journal of Interactive marketing, 21(2), 63-79. Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook “friends:” Social capital and college students’ use of online social network sites. Journal of ComputerMediated Communication, 12(4), 1143–1168. 121 Festinger, L. (1957). Cognitive dissonance theory. Primary Prevention of HIV/AIDS: Psychological Approaches. Newbury Park, California, Sage Publications. Field A. (2005). Discovering statistics using SPSS. Second edition. Sage. Fishbein, M. E. (1967). Readings in attitude theory and measurement. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of marketing research, 39-50. Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149-1160. George, D. (2011). SPSS for windows step by step: A simple study guide and reference, 17.0 update, 10/e. Pearson Education. India. George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and reference. 11.0 update (4th ed.). Boston: Allyn & Bacon. Gibson, E. J. (2003). The world is so full of a number of things: On specification and perceptual learning. Ecological psychology, 15(4), 283-287. Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin. Golbeck, J., Robles, C., Edmondson, M., & Turner, K. (2011, October). Predicting personality from Twitter. In Privacy, Security, Risk and Trust (PASSAT) and 2011 IEEE Third International Conference on Social Computing (SocialCom), 2011 IEEE Third International Conference on (pp. 149–156). IEEE. Guadagnoli, E., & Velicer, W. F. (1988). Relation to sample size to the stability of component patterns. Psychological Bulletin, 103(2), 265-275. Hair, J.F., Black, W.C., Babin, B.J., & Anderson, R.E. (2010). Multivariate Data Analysis. Seventh Edition. Prentice Hall, Upper Saddle River, New Jersey. Hervet, G., Guérard, K., Tremblay, S., & Chtourou, M. S. (2011). Is banner blindness genuine? Eye tracking internet text advertising. Applied cognitive psychology, 25(5), 708-716. Hinz, O., Skiera, B., Barrot, C., & Becker, J. U. (2011). Seeding strategies for viral marketing: An empirical comparison. Journal of Marketing, 75(6), 55-71. Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand engagement in social media: Conceptualization, scale development and validation. Journal of interactive marketing, 28(2), 149-165. 122 Hoffman, D. L., & Novak, T. P. (2011). “Social media strategy.” In V. Shankar & G. S. Carpenter (Eds.), Handbook on Marketing Strategy. Edward Elgar Publishing, Ltd. Northampton, Massachusetts. Hoffman, D. L., & Novak, T. P. (2012). Why do people use social media? Empirical findings and a new theoretical framework for social media goal pursuit. Empirical Findings and a New Theoretical Framework for Social Media Goal Pursuit (January 17, 2012). Hoffman, D. L., Novak, T. P., & Kang, H. (2016). Let’s get closer: Feelings of connectedness from using social media with implications for brand outcomes. Available at SSRN. Hoy, M. G., & Milne, G. (2010). Gender differences in privacy-related measures for young adult Facebook users. Journal of Interactive Advertising, 10(2), 28-45. Hughes, D. J., Rowe, M., Batey, M., & Lee, A. (2012). A tale of two sites: Twitter vs. Facebook and the personality predictors of social media usage. Computers in Human Behavior, 28(2), 561–569. Hurrle, D., & Postatny, J. (2015). Social media for scientific institutions: How to attract young academics by using social media as a marketing tool. Berlin, Germany: Springer. Instagram (2015, May 2nd). Retrieved from https://www.instagram.com/press/?hl=en Jackson, C. M., Chow, S., & Leitch, R. A. (1997). Toward an understanding of the behavioral intention to use an information system. Decision sciences, 28(2), 357-389. Jansen, B. J., Zhang, M., Sobel, K., & Chowdury, A. (2009). Twitter power: Tweets as electronic word of mouth. Journal of the American society for information science and technology, 60(11), 2169-2188. Jeong, Y., & Coyle, E. (2014). What are you worrying about on facebook and twitter? an empirical investigation of young social network site users’ privacy perceptions and behaviors. Journal of Interactive Advertising, 14(2), 51-59. Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of social media. Business Horizons, 53(1), 59–68. Keenan, A., & Shiri, A. (2009). Sociability and social interaction on social networking websites. Library Review, 58(6), 438–450. Kelly, L., Kerr, G., & Drennan, J. (2010). Avoidance of advertising in social networking sites: The teenage perspective. Journal of Interactive Advertising, 10(2), 16-27. Kim, Y., & Yoon, H. J. (2014). What makes people" like" comedic-violence advertisements? A model for predicting attitude and sharing intention. Journal of Advertising Research, 54(2), 217-232. 123 Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development. New Jersey: Prentice-Hall. Kwak, H., Lee, C., Park, H., & Moon, S. (2010, April). What is Twitter, a social network or a news media?. In Proceedings of the 19th international conference on World wide web (pp. 591-600). ACM. Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46-70. Lasswell, Harold D. (1948), "The Structure and Function of Communication in Society," in The Process and Effects of Mass Communication, Wilbur Schramm and Donald F. Roberts, eds., Urbana: University of Illinois Press, 1971, 84-99. Lee, Y. J., & Ahn, H. Y. (2013). Interaction effects of perceived sponsor motives and Facebook credibility on willingness to visit social cause Facebook page. Journal of Interactive Advertising, 13(1), 41-52. Lee, C. S., & Ma, L. (2012). News sharing in social media: The effect of gratifications and prior experience. Computers in Human Behavior, 28(2), 331–339. Leung, L. (2001). "College student motives for chatting on ICQ". New Media & Society. 3 (4): 482–500. Li, H., Edwards, S. M., & Lee, J. H. (2002). Measuring the intrusiveness of advertisements: Scale development and validation. Journal of Advertising, 31(2), 37-47. Logan, K. (2014). Why isn't everyone doing it? A Comparison of antecedents to following brands on Twitter and Facebook. Journal of Interactive Advertising, 14(2), 60-72. Mabry, E., & Porter, L. (2010). Movies and Myspace: The effectiveness of official Web sites versus online promotional contests. Journal of Interactive Advertising, 10(2), 1-15. McLuhan, M., & Fiore, Q. (1967). The medium is the message. New York, 123, 126-128. McGrenere, J., & Ho, W. (2000, May). Affordances: Clarifying and evolving a concept. In Graphics Interface (Vol. 2000, pp. 179–186). McGuire, W. J. (1969). The nature of attitudes and attitude change. The Handbook of Social Psychology, 3(2), 136-314. Moon, J. W., & Kim, Y. G. (2001). Extending the TAM for a World-Wide-Web context. Information & Management, 38(4), 217-230. 124 Morrison, M. A., Cheong, H. J., & McMillan, S. J. (2013). Posting, Lurking, and Networking: Behaviors and Characteristics of Consumers in the Context of User-Generated Content. Journal of Interactive Advertising, 13(2), 97-108. Moore, K., & McElroy, J. C. (2012). The influence of personality on Facebook usage, wall postings, and regret. Computers in Human Behavior, 28(1), 267-274. Muk, A., & Chung, C. (2014). Driving consumers to become fans of brand pages: A theoretical framework. Journal of Interactive Advertising, 14(1), 1-10. Muntinga, D. G., Moorman, M., & Smit, E. G. (2011). Introducing COBRAs: Exploring motivations for brand-related social media use. International Journal of Advertising, 30(1), 13–46. Nagy, J., & Midha, A. (2014). The value of earned audiences: how social interactions amplify TV impact. Journal of Advertising Research, 54(4), 448-453. Nicovich, S. G. (2005). The effect of involvement on ad judgment in a video game environment: The mediating role of presence. Journal of Interactive Advertising, 6(1), 29-39. Norman, D. A. (1988). The psychology of everyday things. New York: Basic Books. Norman, D. A. (1999). Affordance, conventions, and design. Interactions, 6(3), 38–43. Norman, D. A. (2008). THE WAY I SEE IT Signifiers, not affordances. Interactions, 15(6), 18– 19. Norman, D. A. (2013). The design of everyday things: Revised and expanded edition. New York: Basic Books. Nunnally, J. C. & Bernstein, I. H. (1994). Psychometric Theory (3rd ed.). New York: McGrawHill. Nunnally, J. C., Bernstein, I. H., & Berge, J. M. T. (1967). Psychometric Theory (Vol. 226). New York: McGraw-Hill. Orchard, L. J., Fullwood, C., Galbraith, N., & Morris, N. (2014). Individual differences as predictors of social networking. Journal of Computer-Mediated Communication, 19(3), 388-402. Paek, H. J., Hove, T., Ju Jeong, H., & Kim, M. (2011). Peer or expert? The persuasive impact of YouTube public service announcement producers. International Journal of Advertising, 30(1), 161-188. 125 Pan, L. Y., & Chiou, J. S. (2011). How much can you trust online information? Cues for perceived trustworthiness of consumer-generated online information. Journal of Interactive Marketing, 25(2), 67-74. Park, N., Kee, K. F., & Valenzuela, S. (2009). Being immersed in social networking environment: Facebook groups, uses and gratifications, and social outcomes. CyberPsychology & Behavior, 12(6), 729–733. Peters, K., Chen, Y., Kaplan, A. M., Ognibeni, B., & Pauwels, K. (2013). Social media metrics—A framework and guidelines for managing social media. Journal of Interactive Marketing, 27(4), 281-298. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, 123-205. Pew Research (2017, January 12). Social media fact sheet. Retrieved from http://www.pewinternet.org/fact-sheet/social-media/ Pfeiffer, M., & Zinnbauer, M. (2010). Can old media enhance new media? Journal of Advertising Research, 50(1), 42-49. Prendergast, G., Ko, D., & Siu Yin, V. Y. (2010). Online word of mouth and consumer purchase intentions. International Journal of Advertising, 29(5), 687-708. Pruessner, J. C., Kirschbaum, C., Meinlschmid, G., & Hellhammer, D. H. (2003). Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology, 28(7), 916-931. Quan-Haase, A., & Young, A. L. (2010). Uses and gratifications of social media: A comparison of Facebook and instant messaging. Bulletin of Science, Technology & Society, 30(5), 350– 361. Raacke, John; Jennifer Bonds-Raacke (2008). "MySpace and Facebook: Applying the Uses and Gratifications Theory to Exploring Friend-Networking Sites". Cyber Psychology & Behavior. 11 (2): 169–174. Rodgers, S., & Thorson, E. (2000). The interactive advertising model: How users perceive and process online ads. Journal of Interactive Advertising, 1(1), 41-60. Rubin, A. M. (1984). Ritualized and instrumental television viewing. Journal of Communication, 34(3), 67-77. Schau, H. J., Muñiz Jr, A. M., & Arnould, E. J. (2009). How brand community practices create value. Journal of Marketing, 73(5), 30-51. 126 Schivinski, B., Christodoulides, G., & Dabrowski, D. (2016). Measuring Consumers' Engagement With Brand-Related Social-Media Content. Journal of Advertising Research, 56(1), 64-80. Schrammel, J., Köffel, C., & Tscheligi, M. (2009, September). Personality traits, usage patterns and information disclosure in online communities. In Proceedings of the 23rd British HCI Group annual conference on people and computers: Celebrating people and technology (pp. 169–174). British Computer Society. Seidman, G. (2013). Self-presentation and belonging on Facebook: How personality influences social media use and motivations. Personality and Individual Differences, 54(3), 402–407. Selfhout, M., Burk, W., Branje, S., Denissen, J., Van Aken, M., & Meeus, W. (2010). Emerging late adolescent friendship networks and big five personality traits: A social network approach. Journal of Personality, 78(2), 509–538. Seraj, M. (2012). We create, we connect, we respect, therefore we are: intellectual, social, and cultural value in online communities. Journal of Interactive Marketing, 26(4), 209-222. Shan, Y., & King, K. W. (2015). The effects of interpersonal tie strength and subjective norms on consumers' brand-related eWOM referral intentions. Journal of Interactive Advertising, 15(1), 16-27. Shao, G. (2009). Understanding the appeal of user-generated media: A uses and gratification perspective. Internet Research, 19(1), 7–25. Smith, T., Coyle, J. R., Lightfoot, E., & Scott, A. (2007). Reconsidering models of influence: the relationship between consumer social networks and word-of-mouth effectiveness. Journal of Advertising Research, 47(4), 387-397. Smith, A. N., Fischer, E., & Yongjian, C. (2012). How does brand-related user-generated content differ across YouTube, Facebook, and Twitter? Journal of Interactive Marketing, 26(2), 102-113. Sook Kwon, E., Kim, E., Sung, Y., & Yun Yoo, C. (2014). Brand followers: Consumer motivation and attitude towards brand communications on Twitter. International Journal of Advertising, 33(4), 657-680. Speck, P. S., & Elliott, M. T. (1997). Predictors of advertising avoidance in print and broadcast media. Journal of Advertising, 26(3), 61-76. Spotts, H. E., Purvis, S. C., & Patnaik, S. (2014). How digital conversations reinforce Super Bowl advertising. Journal of Advertising Research, 54(4), 454-468. 127 Statista (2016, May 2nd). Leading social media websites in the United States in February 2016, based on share of visits. Retrieved from http://www.statista.com/statistics/265773/marketshare-of-the-most-popular-social-media-websites-in-the-us/ Stern, B. B. (1994). A revised communication model for advertising: Multiple dimensions of the source, the message, and the recipient. Journal of Advertising, 23(2), 5-15. Stewart, D. W., & Ward, S. (1994). Media effects on advertising. Media effects: Advances in theory and research, 315-363. Stewart, D. W., Pavlou, P., & Ward, S. (2002). Media influences on marketing communications. Media effects: Advances in theory and research, 353-396. Tabachnick BG and Fidell LS (2007) Using multivariate statistics. Fifth Edition. Pearson Education Inc Thompson, D. V., & Malaviya, P. (2013). Consumer-generated ads: does awareness of advertising co-creation help or hurt persuasion? Journal of Marketing, 77(3), 33-47. Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at a time. Human Communication Research, 28(3), 317–348. Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word-of-mouth versus traditional marketing: findings from an internet social networking site. Journal of Marketing, 73(5), 90-102. Tsai, W. H. S., & Men, L. R. (2013). Motivations and antecedents of consumer engagement with brand pages on social networking sites. Journal of Interactive Advertising, 13(2), 76-87. Tsang, M. M., Ho, S. C., & Liang, T. P. (2004). Consumer attitudes toward mobile advertising: An empirical study. International Journal of Electronic Commerce, 8(3), 65-78. VanMeter, R. A., Grisaffe, D. B., & Chonko, L. B. (2015). Of “Likes” and “Pins”: The effects of consumers' attachment to social media. Journal of Interactive Marketing, 32, 70-88. Van Noort, G., Antheunis, M. L., & Van Reijmersdal, E. A. (2012). Social connections and the persuasiveness of viral campaigns in social network sites: Persuasive intent as the underlying mechanism. Journal of Marketing Communications, 18(1), 39-53. van Reijmersdal, E. A., Rozendaal, E., Smink, N., van Noort, G., & Buijzen, M. (2017). Processes and effects of targeted online advertising among children. International Journal of Advertising, 36(3), 396-414. Van-Tien Dao, W., Nhat Hanh Le, A., Ming-Sung Cheng, J., & Chao Chen, D. (2014). Social media advertising value: The case of transitional economies in Southeast Asia. International Journal of Advertising, 33(2), 271-294. 128 Viswanath, B., Mislove, A., Cha, M., & Gummadi, K. P. (2009, August). On the evolution of user interaction in Facebook. In Proceedings of the 2nd ACM workshop on online social networks (pp. 37–42). ACM. Vogt, W. P., Gardner, D. C., Haeffele, L. M., & Vogt, E. R. (2014). Selecting the right analyses for your data: Quantitative, qualitative, and mixed methods. Guilford Publications. Voorhees, C. M., Brady, M. K., Calantone, R., & Ramirez, E. (2016). Discriminant validity testing in marketing: an analysis, causes for concern, and proposed remedies. Journal of the Academy of Marketing Science,44(1), 119-134. Wang, A. (2006). Advertising engagement: A driver of message involvement on message effects. Journal of Advertising Research, 46(4), 355-368. Wang, Zheng; John M. Tchernev; Tyler Solloway (2012). "A Dynamic Longitudinal Examination of Social Media Use, Needs, and Gratifications Among College Students". Computers in Human Behavior, 28(5), 1829–1839. Whiting, A., & Williams, D. (2013). Why people use social media: A uses and gratifications approach. Qualitative Market Research: An International Journal, 16(4), 362–369. Woodside, A.G. and Soni, P.K. (1990) ``Performance analysis of advertising in competing media vehicles'', Journal of Advertising Research, Vol. 30, February/March, pp. 53-66. Wright, P. L. (1974). Analyzing media effects on advertising responses. Public Opinion Quarterly, 38(2), 192-205. Wu, L. (2016). Understanding the Impact of Media Engagement on the Perceived Value and Acceptance of Advertising Within Mobile Social Networks. Journal of Interactive Advertising, 16(1), 59-73. Yeh, Y. H., & Choi, S. M. (2011). MINI-lovers, maxi-mouths: An investigation of antecedents to eWOM intention among brand community members. Journal of Marketing Communications, 17(3), 145-162. Yi, Y. (1990). The effects of contextual priming in print advertisements. Journal of Consumer Research, 17(2), 215-222. Zhang, J., & Mao, E. (2008). Understanding the acceptance of mobile SMS advertising among young Chinese consumers. Psychology & Marketing, 25(8), 787-805. Zhao, Y., Liu, J., Tang, J., & Zhu, Q. (2013, March). Conceptualizing perceived affordances in social media interaction design. In Aslib Proceedings (Vol. 65, No. 3, pp. 289–303). Emerald Group Publishing Limited. 129