Why Do People Use Social Media? Empirical Findings and a ... - SSRN

Why Do People Use Social Media?

Why Do People Use Social Media? Empirical Findings and a New Theoretical Framework for Social Media Goal Pursuit

Donna L. Hoffman Albert O. Steffey Professor of Marketing UC Riverside [email protected]

Thomas P. Novak Albert O. Steffey Professor of Marketing UC Riverside [email protected]

This research was supported by the National Science Foundation (Grant # IIS-1114828).

Electronic copy available at: http://ssrn.com/abstract=1989586

1


2

Why Do People Use Social Media? Empirical Findings and a New Theoretical Framework for Social Media Goal Pursuit

Abstract Why do people use social media? We examine how and why people use social media in the context of their basic needs for autonomy, competence and relatedness, intrinsic and external motivations, and well-being perceptions. Results show that motivations differentially drive social media goal pursuit, and users with different primary social media goals differ in perceptions of well-being. Using these results, we develop a testable theoretical framework for social media goals defined by two higher-order dimensions that contrast the primary focus of the online interaction with the primary direction of the online interaction. The framework may be useful to further understanding of the relationship between users’ social media behaviors and subjective well-being in the context of their fundamental needs and motives.

Keywords: social media, online interaction, interaction focus, interaction direction, selfdetermination theory, well-being

Electronic copy available at: http://ssrn.com/abstract=1989586


3

Why Do People Use Social Media? Empirical Findings and a New Theoretical Framework for Social Media Goal Pursuit As the Internet continues its inexorable march toward ubiquity, many are wondering what consequences for our society’s well-being lie in its path. The question is becoming more important as online social networking and “social sharing” behaviors increase and as “alwayson” Internet use becomes more personal, being accessed as much or more on mobile devices than it is on the PC (Meeker 2010). In late 2010, it was widely reported that visits to Facebook now account for 25 percent of page views and ten percent of all Internet visits in the United States (Dougherty 2010), with Facebook surpassing Google in share of visits (Dougherty 2010). Internet users in the United States now spend more than one-quarter of their online time on social media applications like blogs and social networking sites (Nielsen Blog 2010). Google’s CEO recently announced that thirty-five hours of video are uploaded to YouTube every minute (Schmidt 2010), more than double the amount just two years ago. New social search applications like Quora, which facilitates answers to questions from one’s online friends in real-time, suggests that social media are likely to change the dynamics of organic search. Zynga, the social gaming company known for games like Mafia Wars and FarmVille, has more than 320 million users with 3 million concurrent users during peak periods (Chiang 2010) and American Express points can now be redeemed for limited editions of virtual goods to be used in Zynga games (Poletti 2010). Social commerce applications like Beautylish and Lockerz, along with social coupon sites like Groupon and Living Social are becoming increasingly popular as they incorporate user-generated content and group interaction to support shopping activities. Social networking continues to evolve with “personal networks” like Path, Gravity, Aro and GroupMe, distributed social browsing applications like Meebo and social sharing


4

organized around photos with applications like Instagram, Picplz, Twitpic and Dailybooth. As social media applications continue to proliferate and the dynamics of online social interaction continue to evolve, communication researchers are seeking to develop a deeper understanding of how and why people use social media so that theoretically consistent models linking user motivations, social media goals, and perceptions of well-being can be constructed (Hoffman 2012). The need for such models is increasingly acute as user participation in new forms of social media introduces a potential for both positive and negative outcomes on well-being. Social media allow people to connect – and re-connect – sometimes over large distances, offer opportunities for self-expression that may not always be possible offline, provide opportunities for learning and information sharing that are unprecedented, and support users’ needs to control their online experiences. Zynga’s FrontierVille game has hosted 650,000 same-sex marriages in the six months since the game went live (Chiang 2010), possibly more than have occurred in the real world. On the other hand, location-based check-in services, for example, provide opportunities to intrude on user’s privacy and safety. In early 2010, the website pleaserobme.com attracted considerable media attention by showcasing check-in updates from foursquare users who had indicated they were currently at a location other than their home, with the implication that their home was currently unprotected and represented an attractive burglary target (Siegler 2010). Despite some negative concerns, people are primarily attracted to social media for the broad range of goals that can be adopted. Hoffman and Novak (2011) argue that the fundamental interactivity of social media allows for four higher-order goals: connect, create, consume and control. Thus, social media enable and facilitate interactions that connect people. These social


5

media conversations occur through web- or mobile-based applications that people use to create (i.e. post, upload, blog) and consume (i.e. read, watch, listen to) content. Finally, social media applications give individuals a greater ability to manage their reputations and control the applications (e.g. page layout, tagging, rating) and online “settings” such as profile and privacy options. These “4Cs” capabilities of social media undoubtedly explain in part why so many people spend so much of their time using social media and why social media are so popular. Not surprisingly, this popularity has stimulated a surge of research in recent years examining social media usage. Research has identified literally hundreds of motivations underlying why people use social media in a wide variety of contexts including blogs, Twitter, virtual worlds, YouTube and many others (Novak 2008). There has also been a great deal of descriptive research examining different social media behaviors. Yet despite the plethora of research on social media usage, there is little unifying theory to guide analysis. This paper represents a first step on this path toward the development of a unifying theory. We examine two research questions in two large-scale studies and then use the results of these studies to formulate a testable theoretical framework for evaluating social media goal pursuit. Our first research question examines social media goal pursuit in the context of the core social motive of the need to relate (Fiske 2004), the additional fundamental needs for autonomy and competence (Deci and Ryan 1985), individuals’ orientation toward intrinsic and external locus of causality (Deci and Ryan 2000) and two aspects of self-esteem in the context of social identity (Luhtanen and Crocker 1992). Our second research question evaluates the relationship between social media goals and perceptions of well-being (Diener, et.al 1985; Diener, et.al. 2009). The 4Cs of connecting, creating, consuming and controlling social media experiences (Hoffman and Novak 2011) serve as our initial conceptual framework for evaluating how user


6

motivations may differentially drive social media goal pursuit and how those goals and several measures of subjective well-being are associated. The paper is organized as follows. First, we briefly summarize related work and then present our motivational framework and key research questions. In the third and fourth sections we present the results from two large-scale studies. In section five, we draw on these results to develop a theoretical framework that can support future inquiry.

RELEVANT LITERATURE Social Media Motivations and Goal Pursuit Social media, which by definition involve individuals relating to each other, represent new and fertile ground for exploring what drives online social behavior. Not surprisingly, considerable research has addressed the question of why people use social media in varied contexts. Literally hundreds of motivations underlying why people use social media in a wide variety of contexts including blogs, Twitter, virtual worlds, YouTube and many others have been identified (Novak 2008). In a review of more than 30 articles that examined over 100 objectives, goals or motivations for using social media, Novak (2008) uncovered no fewer than 22 distinct motivational categories of social media use (see Table 1). These include, for example, achievement (Kuznetsov 2006), information (Weiss, Lurie, Macinnis 2008), peer pressure (Jung, Youn and McClung 2007), positive experience (Hoffman and Novak 2009), self understanding (Zhao, Grasmuck and Martin 2008) and social interaction (Lenhardt and Fox 2006), among others. While it is clear that people use social media for many different reasons, the lack of a unifying framework hinders deeper understanding of the fundamental motivations driving social media use.


7

---Insert Table 1 about here--Positive Outcomes From Social Connections Recent research has uncovered a strong positive connection between social connections and positive outcomes. For example, the more socially active seniors are, the more likely they are to enjoy improved cognitive and motor functioning, well into their old age (Yaffe, et al. 2009). Another recent study found that the more time seniors spent socializing, the less motor functions declined (Buchman, et al. 2009). Intriguingly, one study has documented a relationship between the size of a person’s social network and their perception of pain: patients with smaller networks report more pain and patients with larger networks report less pain (Mitchinson, et. al 2008). This finding is interesting because it suggests that larger social networks, regardless of the quality of connections, can have positive influences on well-being. Jetten, et al. (2009) reviewed a number of studies in this domain and concluded that individuals who are members of many diverse social networks are more resilient and experience more mental and physical well being compared to individuals who are not as socially connected. Similar positive benefits may also accrue from online social connections. As the Internet continues to diffuse widely through our society, individuals find it easier to keep in touch with their acquaintances (what Granovetter (1973) calls “weak ties”) and use the Internet to strengthen the bonds underlying their close friends (“strong ties”). Social media also offer people opportunities to express themselves and connect with others even if they are able to derive benefit from such opportunities offline. Collins and Wellman (2010) observed that the more people use the Internet, the more connected they are both online and offline. Bessière, et al. (2008) found that Internet users motivated by the need to communicate with friends and family had lower depression scores than individuals motivated to use the Internet to meet new people and chat online. Wellman, et al. (2001) found that online participation actually supplemented


8

face-to-face and telephone interactions by extending another avenue for communication, although the heaviest users did not reap the benefits to the same degree. Benefits of online communication appear to extend to new relationships as well. McKenna, Green and Gleason (2002) found that participants who met each other for the first time through online chat liked each other more based upon that interaction than did participants who met face-to-face for the first time, and related research found that individuals interacting through online chat are better able to express their true self than individuals in face-to-face interaction (Bargh, McKenna, & Fitzsimons, 2002). Thus, online networks are seen as a beneficial tool that compliments the tools users already use to maintain their interconnectedness. As social networks proliferate, communication researchers are turning their attention to the psychological benefits of online participation. For example, Facebook can help users maintain their close relationships at the same time that it helps build and maintain weak ties among large groups of distant friends and acquaintances, leading to an increased sense of selfesteem in new college students (Ellison, Steinfield, and Lampe 2007). Novak (2012) has found that virtual worlds as a social environment offer the potential as an intervention to improve wellbeing, at least in part by substituting for physical consumption and physical presence. While studies seem to suggest that Internet use is likely to be positively associated with subjective well-being, it is not clear under what conditions we might expect the same beneficial effects for social media. It does seem reasonable that online use might generally be expected to contribute to well-being because it satisfies basic needs for relatedness. Yet, because the impact of Internet use on a person’s well being is likely to be a function of the goals people have for interacting (Bargh and McKenna 2004), a full understanding of the impact of social media use on well-being will necessarily require a detailed motivational account.


9

ANTECEDENTS AND CONSEQUENCES OF SOCIAL MEDIAL GOALS Basic Need Satisfaction. We contend that social media use offers individuals the opportunity to satisfy basic psychological needs. Whether or not these needs are satisfied depends on the goals individuals pursue while engaging in social media behavior and the motivations underlying this goal pursuit. Ultimately, whether individuals experience positive outcomes related to psychological well-being and physical health depend on this motivational structure. Consistent with Fiske’s (2004) delineation of core social motives and Baumeister and Leary’s (1995) hypothesis that the need to belong is a fundamental motivation, we argue that individuals use social media because they are fundamentally motivated to connect with each other and that using social media has the potential to provide these connections. In addition to this need for relatedness, we consider that individuals seek to satisfy needs for autonomy and competence (Deci and Ryan 1985). Autonomy refers to an internal perceived locus of control (de Charms 1968) or a sense that one’s behavior is self-determined, while competence describes a sense of self-efficacy where one has the capabilities to meet the challenges presented by the task. Tamborini, et. al. (2010) have recently shown that media enjoyment results from the satisfaction of fundamental needs. Thus, need satisfaction should have important implications for how people experience social media use. Individual Motivational Orientation. According to self-determination theory, the reasons that people are motivated to engage in a particular behavior are governed by their motivational orientation (Deci and Ryan 2000). To the extent that behaviors are internalized or intrinsically motivated, individuals are expected to exhibit greater enjoyment, satisfaction and general wellbeing from those behaviors (see, for example, Ryan and Deci 2000). Intrinsic motivation for a


10

behavior is supported when the basic psychological needs of competence and autonomy experienced during that behavior are satisfied. Behaviors that are extrinsically motivated can vary in the degree of competence and autonomy they satisfy. Because a perceived internal locus of causality appears so strongly implicated in positive subjective experiences, the extent to which extrinsically motivated behaviors are internalized will also relate to positive subjective experiences like well-being. Ryan and Deci (2000) suggest that the need for relatedness, along with competence and autonomy, can aid in internalizing behaviors that are often extrinsically motivated. Although a considerable amount of research has investigated the role of autonomy and competence in motivating behavior, much less is known about the impact of relatedness needs (Pittman and Zeigler 2007). Individuals for whom social media satisfies these needs should experience a positive impact on well-being (Sheldon, Elliot, Kim, & Kasser, 2001). This line of reasoning is closely related to the idea of optimal online experience conceptualized by Hoffman and Novak (1996). Individuals who experience flow during their online navigational experiences are more likely to achieve positive outcomes compared to individuals who cannot attain these compelling online experiences (Novak, Hoffman and Yung 2000). Collective Self-Esteem. Recently, Leary (2007) has argued that self-esteem serves as a measure or sociometer of an individual’s relational value to other people. This motivational perspective suggests that self-esteem helps individuals increase their social value and is not pursued for its intrinsic value. Because social media are, by definition, social, it stands to reason that self-esteem may be an important predictor of social media goal pursuit. We argue that the relevant concept of self-esteem is a collective one that captures an individual’s evaluation of their self-esteem in the context of their social identity (Luhtanen and Crocker 1992). Two


11

components of collective self-esteem are highly likely to motivate goal pursuit: private collective self-esteem and importance to identity collective self-esteem. Private self-esteem taps an individual’s judgment of the value they place on their social groups and identity self-esteem captures the importance of these social groups to their self-concept. We theorize that these components of collective self-esteem motivate users to pursue a positive online social identity among the people in their online social groups and differentially predict social media goal pursuit. Overview of Studies 1 and 2 We designed two large-scale studies to address the primary research questions of how and why people use social media in the context of their needs, motivations, and perceptions of well-being. In study 1 we elicit consumers’ primary goals for using social media and develop a goal typology based upon these elicited goals. We expect that the 4Cs (connect, create, consume, and control) will explain a substantial proportion of the variation among consumer’s social media goals. We also determine how well the pursuit of each of the 4C’s can be predicted from and are theoretically consistent with constructs identified in our literature review - goal-level motivations, individual-level needs/motivations, and related social media characteristics. In study 2, we develop specific goal statements based upon the goals identified in study 1 and ask people to tell us which of these social media goals makes them happy. Our primary interest in study 2 is to evaluate whether individuals with different primary 4Cs social media goals differ in terms of their satisfaction with life as a whole, happiness with social media in general and the degree to which pursuing specific social media goals makes them happy.


12

STUDY ONE Participants. A random sample of 1400 English-speaking participants from a university operated online panel were emailed invitations to a web-based survey. A total of 340 participants (24% of those invited) completed all questions. Demographics of those completing the survey closely mirrored that of the sampling frame of invited panelists. Of 1400 invited panelists, 70% were female, 52% completed at least a college education, and 57% were over 40 years of age. In the final sample of 340 panelists, 72% were female, 56% completed at least a college education, and 59% were over 40 years of age. Procedure. Participants rated their involvement and knowledge with social media, and provided brief written descriptions of their top five specific goals when using social media. Participants then rated the degree to which each of their top five goals was: 1) relevant to connect, create, consume and control goals (based on descriptions derived from our theoretical constructs), 2) related to autonomy, competence and relatedness needs (derived from Sheldon, et al. 2001), and 3) related to intrinsic and external locus of causality motivations (derived from Sheldon and Elliot 2000 and Ryan and Connell 1989). We also measured two aspects of respondents’ social identity in the context of their social media use: private collective self-esteem and importance to identity collective self-esteem (scales adapted from Crocker et. al.1994). Analysis and Results: Goal Coding. A total of 1700 social media goal verbatims were obtained from the 340 participants. These were classified into a smaller set of 27 more general social media goals to more clearly understand how different types of social media goals correspond to the four broad underlying goals of connect, create, consume and control. Figure 1 plots the group centroids for these 27 goal types, based upon a canonical discriminant analysis predicting social media goals from the 4Cs ratings (connect, create, consume, control) of each of


13

the 1700 goals, with the 27 goals serving as the groups. The first two dimensions plotted in Figure 1 account for 78% of the variance among the 27 social media goals, showing that the 4Cs provide a good higher-level explanation of specific social media goals. Differences among the 27 social media goals in terms of consume, connect and create goals are well represented in the plot. Control less clearly discriminates among the 27 specific goals, suggesting that control “cuts across” goals and relates to how a given goal is approached. ---Insert Figure 1 about here--Analysis and Results: Multilevel Model Predicting Goals from Needs and Motivations. The 27 social media goal types were further classified as either social (49.8% of goals) or nonsocial (50.2% of goals). Examples of social goals were connecting with friends and family, using social media to meet new people, sharing pictures and videos with friends, and reconnecting with people one had lost touch with. Non-social goals included reading the news, learning about popular events, listening to music, downloading videos, finding deals, and researching products. The data collected in study 1 are multilevel, in that five social media goals are nested within 340 individuals. The set of five goals elicited for each person may differ considerably from person to person, with individual level heterogeneity leading to correlated errors in goallevel ratings. Thus, multilevel linear models (also known as linear mixed models, hierarchical linear models, or random effects models) were used to account for correlated errors in the multilevel data e.g. Hox 2002; Kreft and deLeeuw 1998). Level 1 variables are those measured at the goal level (1700 observations) while level 2 variables are those measured at the individual level (340 observations). Multilevel models were fit separately to each of four dependent measures (i.e. the 4Cs goal-specific ratings of connect, create, consume, and control). We first fit the null model


14

containing only a random intercept in order to estimate the intra-class correlation, ρ, and to provide baseline level 1 and level 2 residual variances for comparison with our complete model. Intra-class correlations in the null model reported in Table 2 show that the greatest heterogeneity among individuals is in adopting control goals (ρ = .62) and that the least heterogeneity is in adopting connect goals (ρ = .32), with create goals (ρ = .49) and consume goals (ρ = .42) lying in between. Thus, consumers are relatively homogeneous in their pursuit of connect goals, but differ in the extent that they pursue control goals. ---Insert Table 2 about here--The complete model fit to each of the four dependent measures includes a random intercept and the following additional fixed effects: 1) a level 1 binary variable for goal type (social vs. nonsocial); 2) within person mean centered level 1 measures of autonomy, competence, relatedness, intrinsic LOC and external LOC (i.e. means for each individual were subtracted from the goal-specific ratings); 3) between person level 2 individual means (calculated over each person’s five goals) of autonomy, competence, relatedness, intrinsic LOC and external LOC; and 4) dispositional level 2 measures of private collective self-esteem, importance to identity collective self esteem, social media knowledge, and social media involvement, all level 2 predictors measured at the individual level. Note that the use of level 1 (centered goal-specific ratings subtracting the individual means) and level 2 (the reintroduced subtracted means) versions of autonomy, competence, relatedness, intrinsic LOC and external LOC allows independent assessment of within-person and between-person effects for these five measures (e.g. Kreft and deLeeuw 1998). The complete multilevel model allows us to determine whether and how the 4Cs (i.e., connect, create, consume and control goals) can be predicted from needs, motivation, and


15

dispositions, taking into account both within person (level 1) and between person (level 2) variation in the 4Cs. From table 2, the degree of within-person variation that can be explained (i.e. R2within) ranges from .07 for consume to .37 for connect, and the degree of between-person variation that can be explained (i.e. R2between) ranges from .21 for consume to .64 for create. Having established that we can explain a substantial portion of both within and between-person variation, we turn next to understanding what predicts the 4Cs Table 3 presents fixed parameter estimates for the complete model. The results show that the pursuit of connect, create, consume, and control goals are differentially predicted by needs, motivations, and dispositions. When broadly categorized as social or non-social, connect goals are more likely to be characterized as social (B = .54, p < .01), and consume goals as non-social (B = -.33, p