Social Media as a Research Environment

Online social activities are undoubtedly fascinating due to the many ways they seem novel relative to social life in the offline world. Asynchronous, persistent, and multimedia conversations, malleable self-presentation, fluid group boundaries and social ties, and the freedom from the strictures of geographic proximity serve to distinguish blogs, wall posts, video chat, and tweets from the rich, realtime face-to-face interactions that social scientists have studied for decades. Though it is important to study what makes online interaction distinctive, we are struck by the promise for social science research on general principles and patterns of human social behavior. Despite their superficial differences, offline and online behavior are different manifestations, in different contexts, of such general phenomena as the desire for status, admiration and affection, attention toward maximizing utility under certain circumstances, the influence of normative constraints, and so on. More and more, researchers are turning to records of social behavior from blogs and social networking sites due to the availability, scale, and temporal granularity of the data. In many ways, this is an exciting and promising development. Detailed observations of real time social interactions of millions of people present the social sciences with the opportunity to study populations and collectives rather than samples and individuals, study them continuously rather than at intervals, and without introducing observer effects. However, there are also pitfalls. Like any new instrument, Internet-based sources of observational or experimental data need to be calibrated. When online replications of offline studies produce similar results, this raises our confidence in the earlier findings and in the new methods for finding them. When they differ, we can investigate whether previously thought universals are actually so, or they vary under previously unstudied conditions. In the early days and years of online social science, more questions than answers will surely be uncovered. For this special issue, we have selected papers that not only provide empirical results about online behavior in relation to social and psychological phenomena, but also shed light on the implications of these results for online social science research. It is, for example, of paramount importance to understand under what conditions online behavior generalizes to offline behavior, and when idiosyncratic properties of the online environment may instead be responsible for the observed phenomena, making the online–offline generalization untenable. Some papers in this issue address trust and closely related concepts like reputation, self-presentation, and privacy, all of which loom large in the study of online spaces. The ability to control group membership and sanction group members’ deviant behavior affect whether groups will exist in order or descend into chaos. Beliefs about others’ competence and intentions influence whether one will take their advice or purchase their secondhand goods. Several papers address the topic of online trust empirically, two in the context of Facebook. Stieger et al. found that privacy concerns led people to leave Facebook—this raises interesting methodological implications, as these selection effects may result in a Facebook user population that exhibits less privacy concern than the population from which it is drawn. Kuo et al. examine how the Facebook interface in particular facilitates self-presentation, and Edelmann argues that not engaging in self-presentation, or ‘‘lurking,’’ can be an important form of participation. In other contexts, Zhao et al. examined correlates of trust in health communities, and Chih et al. examined reviewer credibility and its effect on purchase decisions. Online spaces have been recognized as a source of fulfillment for a variety of social needs that one can perhaps not find in one’s geographically local area, perhaps due to being a member of a marginalized group, holding a minority opinion, or being distant from one’s social network. Three papers examined social support, emotion, and related concepts. Rui et al. examined the types of social support provided and sought by health organizations on Twitter, while Liu and Yu found that college students were able to find social support on Facebook, but this was dependent on the availability of social support more generally. Finally, Wang et al. investigated Occupy Wall Street discussions on Twitter, and did not observe a relationship between emotion and political discussion. Five papers address topics related to overall methodological issues of Internet-based social science research. Leng takes on the limitations inherent in generalizing from online to offline, and Callaghan et al. observe that ‘‘contrived’’ online environments may stimulate greater levels of self-disclosure, while Konijn et al. demonstrate the feasibility of conducting research on YouTube, and Fu and Chan address the online– offline relationship by observing a case in which sentiment in social media served as a leading indicator of public opinion as measured via a phone survey. Finally, Moreno et al. address the issues facing university IRBs in relation to Internet-based research, including privacy, consent, and confidentiality.