Hollenbaugh (2021). Self-Presentation in Social Media.pdf

OPEN ACCESS Top-Quality Science Peer Reviewed Open Peer Reviewed Review of Communication Research 2021, Vol.9 ISSN: 2255-4165 Self-Presentation in Social Media: Review and Research Opportunities Erin E. Hollenbaugh Kent State University at Stark, USA [email protected] Abstract This paper reviews e...

OPEN ACCESS Top-Quality Science Peer Reviewed Open Peer Reviewed Review of Communication Research 2021, Vol.9 ISSN: 2255-4165 Self-Presentation in Social Media: Review and Research Opportunities Erin E. Hollenbaugh Kent State University at Stark, USA [email protected] Abstract This paper reviews existing research on self-presentation in social media in order to inform future research. Social media offer seemingly limitless opportunities for strategic self-presentation. Informed by existing self-presentation theories, a review of research on self-presentation in social media revealed three significant context and audience variables that were conceptualized in a model. First, three affordances of social media – anonymity, persistence, and visibility – were discussed, as research has revealed the moderating effects of these affordances between self-presentation goal and the self-presentational content shared in social media. For example, one might expect that social media users are more likely to present their actual selves under conditions of less anonymity, more persistence, and more visibility. On the other hand, the freedom associated with more anonymous, less persistent, and less visible social media may lead to idealized self-presentation. The second finding revealed the impact of other-generated content in the form of likes, comments, tags, and shares on social media users’ self-presentation content, mediated by how they choose to manage such content. The third theme concerned the moderating effect of context collapse on the relationship between goals and self-presentation content. The composition of an impression manager’s audience from one platform to the next varies across social media platforms, impacting and often complicating the attainment of self-presentation goals in the midst of merging networks of people. Social media users have adopted varying ways to navigate the complexities of context collapse in their pursuit of self-presentation. Although we have learned much from this body of literature, a more comprehensive theory of self-presentation in the hypermedia age is needed to further advance this area of research. Highlights • Social media provide increased opportunities and challenges to manage self-presentation. • Existing theories of self-presentation should be updated to include variables unique to social media channels in order to apply to the social web. • Self-presentation is moderated by the technological and social affordances of social media platforms, such as anonymity, persistence, and visibility. • Social media users respond to content provided by others in an effort to manage their online image. • The composition of online networks impacts users’ decision-making about self-presentation content. • This review of literature identifies three key variables – affordances, managing content generated by others, and audience composition – that impact self-presentation in social media. • A new model is presented that identifies moderating relationships among variables, drawn from existing literature. Suggested citation: Hollenbaugh, E. E. (2021). Self-Presentation in social media: Review and research opportunities. Review of Communication Research, 9, 80-98. https://doi.org/10.12840/ISSN.2255-4165.027 Keywords: self-presentation; social media; impression management; context collapse; social networking sites. Editor: Giorgio P. De Marchis (Universidad Complutense de Madrid, Spain). Reviewers who acceepted to sign their review: Anneleen Meeus (KULeuven, BE), Johanna M.F. van Oosten (University of Amsterdam, NL), a third reviewer prefers to stay blind. Received: Jan 14th, 2020 Accepted: Sep, 30th, 2020 Prepublished: Dec, 2020 Published: Jan, 2021 Hollenbaugh Content INTRODUCTION.........................................................................................................81 GOALS AND APPROACH...........................................................................................82 Figure 1......................................................................................................................83 SELF-PRESENTATION AND IDENTITY....................................................................84 AFFORDANCES OF SOCIAL MEDIA........................................................................84 Anonymity..................................................................................................................85 Persistence..................................................................................................................86 Visibility.....................................................................................................................86 MANAGING SELF-PRESENTATION CONTENT FROM OTHERS...........................87 AUDIENCES AND CONTEXT COLLAPSE................................................................88 Context collapse..........................................................................................................88 DISCUSSION................................................................................................................90 REFERENCES..............................................................................................................92 APPENDIX. RESEARCH DATABASES.......................................................................96 Introduction 1997). In 1993, the famous New Yorker cartoon canine declared, ‘On the Internet, nobody knows you’re a dog,’ highlighting the freedom afforded by the Internet to carefully structure one’s self-presentation. Turkle (1997) explored anonymous online interactive environments as spaces for playing with identity and trying on new personae. The hyperpersonal perspective connected structural features of the Internet with users’ ability to control their selective selfpresentation (Walther, 1996). Visual and discursive anonymity reduced people’s inhibitions online, resulting in a host of benign and toxic disinhibition (Suler, 2004). Despite the freedom associated with online communication in the late 20th century (Turkle, 1997) and the dawn of the 21st (Suler, 2004), the creation and proliferation of social media since that time has changed the metaphoric game of self-presentation in at least three ways through social media’s features and use. Decreasing anonymity for users, contributions from audiences, and increased context collapse have shaped self-presentation in social media. First of all, instead of using pseudonyms and avatars to represent one’s identity, social networking sites like Facebook have decreased online anonymity through encouraging users to frequently update profile pictures and even requiring its users to register their actual name (see Facebook, 2020). Although specific platforms and tools afford more anonymity (e.g., Yik Yak [Heston & Birnholtz, 2016] and reddit [Leavitt, 2015]), the more It is well accepted that people engage in selective self-presentation by highlighting certain aspects of themselves while downplaying others (Goffman, 1959). Self-presentation is goal-directed and is performed for an audience, and the success of one’s self-presentation is measured by whether or not the audience accepts this performance (Schlenker, 1985). Early theories of self-presentation focused primarily on face-to-face performance of the self, where the audience and context are bounded and more apparent (Goffman, 1959; Leary & Kowalski, 1990; Schlenker, 1985). As humanity became more engrossed in electronic media, theorists such as Meyrowitz (1986) attempted to reinvent traditional self-presentation theories into mediated environments through his second-generation medium theory (see Meyrowitz, 1994). Early theories of self-presentation such as the dramaturgical perspective (Goffman, 1959) and Schlenker’s (1985) theory of self-identification have found new life in the present, as they have been useful to understand self-presentation processes in social media. Among other things, the massification of the Internet brought with it expanded and exciting avenues for selfpresentation. Research exploring multi-user domains, chatrooms, and personal websites gave rise to new theorizing about expression and self-presentation on the Web (Turkle, www.rcommunicationr.org 81 Self-Presentation in Social Media common uses of social media (e.g., Facebook, Instagram) are less anonymous. The prominence of offline friends and family members in social media users’ friends list further remove opportunities for anonymity (e.g., Anderson & Jiang, 2018; Manago et al., 2012). The grounding of one’s online identity in an offline identity and physical space limits the capacity for highly controlled, selective self-presentation online because of these changes in anonymity afforded by some social media sites. Instead, today’s social media space offers limited control, especially given the opportunity that users have to contribute to or even change the impressions of others. The second way that social media has changed self-presentation is by the increased opportunities for others to contribute content that has implications for one’s self-presentation. Actors wishing to present a specific aspect of themselves must also manage such content provided by others. Net- Attributes of the channel itself provide social information that is necessary to read the social situation and make decisions about self-presentation performances (Meyrowitz, 1986). Though Goffman (1959) and other early self-presentation theorists focused on face-to-face contexts for impression work, it is helpful to examine mediated channels as environments in which self-presentation and facework may occur (Meyrowitz, 1986). Studying self-presentation on social media is important because fulfilling this need is a significant predictor of social media attachment and addiction (Chen, 2019). Toward that end, the author conducted a search of the literature on self-presentation and social media, ultimately focusing on the unique contributions of social media contexts in self-presentation content and processes. In particular, research has shown that the affordances of social media channels (i.e., the interaction between attributes of the technology and its typical uses; Evans et al., 2017), worked publics contribute to one another’s impression management through actions such as tagging, commenting, liking, and sharing (Rui & Stefanone, 2013). In accordance with Brunswik’s (1956) lens model, audiences use all of the available cues to build an impression of a person. According to warranting theory, other-generated content may be more influential on observers’ impressions of social media users than the users’ own posts (Walther et al., 2009). In fact, impressions that others make of social media users are influenced by the attractiveness of users’ friends (Walther et al., 2008). While some social networking sites have tools built in to help manage the contributions of others (i.e., tagging, commenting, liking, and sharing), the use of such tools requires active attention to social media that older users may be unwilling to give (see Ongun & Demirag, 2014). Along with decreasing anonymity and contributions of others, a third way in which social media has impacted self-presentation is through increased context collapse within the large and diverse audience of social media. Much like a wedding where otherwise disparate groups collide in a shared social space, social media friend lists typically integrate networks of varying intimacy levels from many aspects of life: work, school, family, friends, and community (Vitak, 2012). The collapsing of contexts, paired with the invisible nature of the audience in social media, makes selective self-presentation especially difficult. Social actors have adopted varying strategies in response to these challenges, which will be further explored in this paper. management of content from others (i.e., ignoring, deleting, or commenting on others’ activity linked to one’s social media profile), and context collapse (i.e., merging various social networks that would otherwise be separate in offline life) are unique variables that should be integrated into prevailing theories and models of self-presentation. Further, a model is provided that illustrates the possible relationships among variables that were yielded from this research, which is described below and evidenced throughout this paper (see Figure 1). This model can be used to understand online selfpresentation research as well as guide future researchers in refining and testing these predictions. Goals and Approach In a world where social media are ubiquitous (see A. Smith & Anderson, 2018) and where the impressions people leave online will follow them throughout their lives, it is important to more systematically examine the elements of self-presentation on social media platforms. Doing so will modernize existing self-presentation theories that have not been fully tested in online spaces. A functional perspective to selfpresentation focuses on the variables affecting social behavior within a larger system. The purpose of this paper is to review the literature of online self-presentation, focusing on social media, in an effort to organize research findings. The author first used a broad search strategy to collect as many papers as possible. The keywords “self-presentation or 82 2021, 9, 80-98 Self-Presentation in Social Media impression management” AND “social media, social networking sites”, or particular social media applications (e.g., “Twitter”) were searched in all EBSCOhost databases during September – December 2017 (see Appendix). “Self-presentation” and the social media terms revealed 44 relevant results, whereas “impression management” yielded an additional 16 papers. While reviewing the 60 papers uncovered during the initial search, special care was taken to organically identify additional publications that were relevant to the project and brought to light in the literature reviews, yielding an additional 13 papers. Through the process of external review by colleagues and others, approximately 20 additional papers and books were added to the set of review materials. Initially, all relevant themes were identified and considered for this project, such as the impact of individual and cultural variables on self-presentation content (e.g., Pearce & Vitak, 2016), as well as self-presentation of celebrities (e.g., lapse”. Combing through these results yielded another seven publications relevant to the review, resulted in a total of approximately 100 papers. These three themes were identified because they were not only dominant in the literature, but they also represent unique characteristics that are distinct among social media when compared to face-to-face self-presentation. A model is proposed to map out the relationships among study variables (see Figure 1). Overall, this model illustrates the moderating effects of social media affordances and perceived audience on the relationship between one’s identity goal and chosen self-presentation content in social media. Prevailing selfpresentation theories based in face-to-face communication suggest that the context and audience are important factors which moderate the effects of self-presentation motive on how to perform one’s identity (e.g., Goffman, 1959). In social media, the context can be measured as affordances of the chosen social media platform, while invisible audiences are L. R. Smith & Sanderson, 2015) and politicians (e.g., Stanyer, 2008) in today’s influencer culture (e.g., Audrezet et al., 2018). Ultimately, the review focused on the three dominant themes impacting self-presentation in social media: affordances, other-generated content, and context collapse. The corpus of research reviewed was continually updated through July 2020 using the same research databases noted above. However, given the more advanced stage of writing, search terms were refined to include three search strategies: “self-presentation” AND “social media” AND (a) “affordances”, (b) “other-generated content”, or (c) “context col- considered by social media users when making decisions about the content they will post. Additionally, when connected audience members tag and comment on one’s content, these posts contribute heavily to one’s self-presentation performance. Social actors must consider how to manage and respond to such content, bearing in mind that ignoring the content is a strategy as well. The proposed relationships are further described with evidence from the literature review. First, a brief overview of the fundamental elements of self-presentation is offered. Figure 1. Predicted Relationships Among Variables in Self-Presentation on Social Media (back to p82) 83 2021, 9, 80-98 Hollenbaugh Self-Presentation and Identity Effectiveness is determined by whether or not the audience accepts one’s self-presentation as accurate. According to self-verification theory, the reactions of others to a performance are important contributors to one’s self-conception (Swann, 1983). People desire others see them the way they see themselves, and social media provide countless opportunities to perform and test self-conceptions for varying audiences. For example, recent research has shown that Facebook users consider paralinguistic digital affordances (i.e., one-click reactions) as one measure of success (Carr et al., 2018). In a qualitative study of adolescents, Barbovschi et al. (2018) found that young girls specifically identified the number of likes on a post as validation of their popularity. Swann (1983) explained that people construct opportunity structures that will confirm their self-conceptions. In social media, this may involve carefully curating a friends’ list that will accept the user’s self-presentation content and respond Through psychological processes of self-identification, people develop private and public images of themselves (Schlenker, 1985). Images of the self, known as self-concept, are multidimensional. For example, the actual self includes characteristics that one possesses, while the ideal self includes characteristics one wishes to have (Higgins, 1987). Self-presentation involves “attempting to control images of self to others” (Schlenker, 1985, p. 67). Many communication cues, verbal and nonverbal, are used by actors to try to influence audiences’ perceptions of one’s identity (Goffman, 1959), whether that is the actual self or ideal self. Self-disclosure, the revelation of personal information (i.e., thoughts or feelings) to others (Derlega et al., 2008), is distinct from self-presentation, yet it is an essential tool for those attempting to present their actual or ideal image (Goffman, 1959). Self-disclosure is a core component of self-presentation, especially online because in the largely text-based environment of social media, verbal disclosures are often highly controlled and audience-specific. Along with verbal self-disclosures, research has found that social media users share pictures (e.g., Barbovschi et al., 2018; Pounders et al., 2016) and location check-ins (e.g., Schwartz & Halegoua, 2015) in hopes of influencing the impressions that others make of them. Central to the topic of self-presentation is the notion of the audience. Without an audience, there is no self-presentation. Social media users heavily weigh the perceptions of their audiences when posting and liking self-presentation content online (Lowe-Calverley & Grieve, 2018). Selective self-presentation varies by audience, such that people may present some aspects of themselves in a certain way to one audience but different parts of their identity in a different way to another (Goffman, 1959). Schlenker (1985) identified three possible audiences who may bear witness to one’s efforts at self-presentation. “Interactants” are the actual people who receive one’s self-presentation messages. “Imagined audiences” are internalized audiences whom one considers when engaging in self-presentation (Schlenker, 1985). For example, one might consider what parents, teachers, or clergy members would think of a message prior to publishing it. Finally, the presenter is their own audience. Social actors bring their own “internalized knowledge and standards for self-regulation” during self-presentation (Schlenker, 1985, p. 66). www.rcommunicationr.org positively to posts. Importantly, the effectiveness of selfpresentation may be used as feedback to influence future impression management. If others’ reactions do not confirm and accept the user’s self-presentation, then the performer may scrutinize that feedback to determine its acceptability. Overall, self-verification theory proposes that social actors will typically find ways to preserve their sense of self by dismissing or reframing disconfirming feedback (Swann, 1983). With a greater understanding of self-presentation messages and processes, attention is now turned to the first theme identified in the literature: the moderating effects of social media affordances on self-presentation. When social media users prepare to present their actual or ideal selves, they must accept that certain features of the chosen medium can impact the messages that they ultimately craft to fulfill their identity goal (see Figure 1). Affordances of Social Media The features of a communication technology help inform people of its use (Norman, 1988). Structural affordances emerge when one examines the relationship between a technology and the ways that humans use it (Evans et al., 2017). This review revealed three affordances that may be significant to self-presentation in social media: (a) anonymity, or the disconnect between one’s offline and online identity, (b) persistence, or the durability of a message to remain available 84 Self-Presentation in Social Media online, and (c) visibility, or the appearance of messages easily in search results or newsfeeds. Although these three structural affordances might influence how people use communication technology to fulfill goals, it does not necessarily determine its use (Evans et al., 2017). For example, although Twitter is highly visible and allows the user to broadcast a message, a person may use Twitter to send messages to a small group of friends, ignoring the public nature of that technology. Examining the affordances of social media channels is one way to consider the moderating effects of the channel on users’ self-presentation efforts. Focusing on affordances also provides one mechanism by which scholars can build upon past research findings amid a swift current of evolving social media channels and features (DeVito et al., 2017). According to Bayer et al. (2020), “an affordance approach may help researchers synthesize findings across studies, time ies have supported the connection between anonymity and self-presentation performance. For example, a survey of almost 200 adults in the U.S. explored the relationships between social media affordances and self-presentation. Tumblr users in particular perceived high amounts of anonymity in this medium and, in turn, less constraint with their self-presentation (DeVito et al., 2017). Reddit is another social media platform that affords high anonymity for its users. Leavitt’s (2015) qualitative study found that reddit users established and used ‘throwaway accounts,’ or alternate identities not tied with their primary identity, in order to post unpopular or controversial opinions anonymously. Additionally, reddit users were less likely to use throwaway accounts under conditions of more perceived anonymity, suggesting that the protections of anonymity were found in this social media platform (Leavitt, 2015). Increased innovations in new technology have complicated the effects of anonymity on goal fulfillment through periods, and specific social media platforms” (p. 475). In an effort to catalog affordances in communication research, Evans et al. (2017) argued for several affordances that are relevant to self-presentation in social media, which also appeared as significant predictors in the reviewed literature: anonymity, persistence, and visibility. These three affordances moderate the relationship between identity goals and self-presentation content (see Figure 1). self-presentation. Location-aware apps, such as Yik Yak, mark their users with a social identity based in geographic location, which can lead to increased use of language that invokes local group identity while still maintaining personal anonymity (Heston & Birnholtz, 2016). Location-based ‘hookup’ apps such as Grindr, designed originally to connect men who have sex with men, introduce unique challenges with managing identity (Birnholtz et al., 2014). Grindr users express a desire to be seen by co-situated others to facilitate casual sex, but often struggle with disclosing their offline identities given perceived stigma (Birnholtz et al., 2014; Blackwell et al., 2014). Qualitative research has shown that Grindr users employ linguistic techniques to distance themselves from potential stigma related to casual sex; for example, using terms such as “fun” and “no strings attached” (“nsa”) were more frequent than specifically naming a “hookup” as the goal (Birnholtz et al., 2014). Another interesting result of this research is that identifiability is determined not only by disclosure of the name and other personal identifiers, but also pictures with the face featured (Blackwell et al., 2014). Taking and posting “selfies,” or pictures taken by and of oneself, has exploded in recent years as a social media practice, due in part to the proliferation of smartphones with cameras (Taylor, 2014). Research has found that females in particular are more likely than males to post selfies that demonstrate positive physical appearance (Jyrkiäinen, 2016; Pounders et al., 2016). While posting a picture of one’s face Anonymity Social media platforms afford anonymity when “the message source is unknown and unspecified” (Scott, 1998, p. 387), which separates a person’s online identity from their offline one. Like all affordances, anonymity exists on a continuum from identified to anonymous (Evans et al., 2017; Scott, 1998). Under conditions of more visual or discursive anonymity, where people are not tied to their offline identities, social media users should feel more uninhibited in their self-presentation (Suler, 2004; Walther, 1996). Walther’s (1996) hyperpersonal perspective argued that anonymity allows for more selective self-presentation. Although the theory is somewhat outdated, the more recent research reviewed below provides support for this prediction. Therefore, we can expect that anonymity moderates the relationship between presentation motives and content, such that increased anonymity allows for presentation that fits our ideal self better. Evidence from quantitative and qualitative research stud- 85 2021, 9, 80-98 Hollenbaugh is in and of itself reducing anonymity, women may post selfies to reflect positively on their offline identity performances. Participants in Pounders et al.’s (2016) qualitative study said they post selfies, in part, to demonstrate that they have the capacity to look good when they want to or on special occasions. For example, one participant said, “…it’s kind of like saying ‘I look nice today, just so you know. Monday through Thursday, um, but check me out on Friday’” (Pounders et al., 2016, p. 1887). Several participants said they liked posting selfies to demonstrate their attractiveness because most people, most of the time do not see them dressed up or with their hair and makeup done. Under conditions of anonymity, online communicators may disclose more information and present themselves in ways that are ideal or preferred, which moderates the effect of self-presentation goal on content (see Figure 1). Research in online dating sites has supported this connection: people across social media challenges, users expressed more concern for the privacy of their information on Instagram than on Snapchat, arguably due to the persistence of content on Instagram (T. R. Choi & Sung, 2018). Snapchat, on the other hand, has less persistence because of the ephemeral or temporary nature of its content. In a survey of over 500 Snapchat and Instagram users, T. R. Choi and Sung (2018) found that Snapchat users were more likely to report revealing their true self, including characteristics they do not feel comfortable displaying offline, due in part to the low persistence of that medium’s use. Regarding the persistence of identity information, DeVito et al. (2017) found that Facebook and LinkedIn were the platforms with the highest persistence of those studied. On the other hand, Tumblr users can easily maintain multiple identities and switch from one self-presentation to another (DeVito et al, 2017). Therefore, one would expect that content shared on Tumblr is less persistent, or less enduring on one’s public persona. Although only a few qualitative and quantitative who anticipate face-to-face encounters with those they meet online (i.e., reduced anonymity) are more likely to be honest in their online dating profiles (e.g., Ellison et al., 2006; Toma et al., 2008). On the other hand, some online daters take advantage of the asynchronicity and reduced nonverbal cues available in social media and share deceptive information about themselves (Ellison et al., 2011). For Ellison et al.’s (2011) participants, this “profile as promise” framework implies that deceptive information shared in online dating apps are realistic and achievable aspects of identity, such as losing weight or earning a promotion at work. It appears that the relationship between anonymity and veracity of selfpresentation content is complicated and in need of further exploration. studies have examined the impact of persistence on selfpresentation specifically, it appears that a relationship does exist. Visibility The use of social media has high visibility when the information can be easily located or retrieved (Evans et al., 2017). Specific features such as the use of trending hashtags in Twitter or posting to a public story in Snapchat suggest increased visibility. Under conditions of high visibility, social media users may be more selective in their self-presentation to convey an ideal image (see Figure 1). DeVito et al. (2017) found that Facebook users perceived this medium to have the highest level of visibility control, such that users can employ built-in tools to manage who can see each piece of content. However, Y. H. Choi and Bazarova (2015) found conflicting results. In their survey of Facebook users, the participants expressed more concern for privacy in Facebook than their private Twitter accounts because of the unbounded nature of the audience where one’s post may be visible to friends of friends, beyond the original intended audience. Velten et al. (2017) argued that Snapchat users typically have high expectations of privacy, given the restrictions on allowing others into the collective boundary. However, when one’s message is ‘screenshot’ and saved, boundary turbulence can occur because a violation of expected invisibility has oc- Persistence Persistence is the durability of a message over time, such that it is archived and available (Evans et al., 2017). It is expected that people using social media channels that have higher persistence will be more deliberate and selective with their self-presentation, due in part to the enduring nature of that content (see Figure 1). Social media research has hailed Facebook as a platform with high persistence; people are often motivated to use Facebook to archive significant life events through posts and photo albums (see Sundar & Limperos, 2013, for a review). Twitter has high content persistence as well, where posts are archived and typically made public (DeVito et al., 2017). When comparing privacy concerns www.rcommunicationr.org 86 Self-Presentation in Social Media Managing Self-Presentation Content from Others curred (Velten et al., 2017). Along with the content and visual components of posts, modern social media users are able to disclose location as an added element of self-presentation, with often increase visibility to those in a geographic area or to those who have also checked into that location. Social media users can communicate spatially through location-based social media networks (e.g., Foursquare), photo-sharing platforms (e.g., Instagram), and mixed-use networks that combine several modes of communication (e.g., Facebook, Twitter; Schwartz & Halegoua, 2015). Although research has focused on the power of these technologies to coordinate our social activities and drive consumerism (see de Souza e Silva, 2013), relatively few studies have examined how the broadcasting of such location-based activity can be employed as a tool for self-presentation (cf. Cramer et al., 2011; Saker, 2017; Schwartz & Halegoua, 2015). Increased visibility may have implications not only for self-presentation, but for safety and well-being as well. For Research reviewed thus far in this paper focuses on the moderating effects of affordances on self-generated content in the form of social media posts. However, the nature of networked publics allows the audience members in the network to contribute to one’s self-presentation as well through comments and images (boyd, 2011). The way that social media users manage content generated by others mediates the effects of that content on the users’ self-presentation content (see Figure 1). Boyd (2011) argued that social media users typically welcome engagement with their networks, but at times this may pose a challenge. Through tagging people in photos, checking others into a location through geotagging, and posting on other people’s timelines, for example, much information may be provided that can impact the impression of another person without permission. In fact, research on warranting theory has shown that other-generated content tends to have a stronger effect on impressions that others create of social media users than the content that users post themselves (e.g., Walther et al., 2009). Social information provided by others is typically judged by observers to have Muslim women living in Copenhagen, social media and physical space are intimately tied together (Waltorp, 2013). The women in Waltorp’s (2013) case study found that social media can be used by powerful others to track their movements and behaviors, potentially fueling stalking behavior. Pearce et al. (2018) identified visibility as the most influential affordance in Azerbaijan, a politically marginalizing culture. Social media users who are highly motivated to present themselves as dissenters must carefully craft their messages to allow for self-expression and facilitate connection with others while also protecting themselves from punishment (Pearce et al., 2018). Though most of the research on visibility as a vehicle for self-presentation demonstrates how social media users actively curate their online identities through managing check-ins, there is a potential “dark side” application of this feature of social media. In short, increased visibility of self-presentation content may lead to more careful posting when revealing their true or ideal self. However, various interaction effects among anonymity, persistence, and visibility should be examined to clarify differing effects in self-presentation. Without a strong body of quantitative results, it is difficult to draw specific hypotheses. higher warranting value, or the value to “draw a reliable connection between a presented persona online and a corporeally anchored person in the physical world” (Walther et al., 2009, p. 232). For example, research has shown that a Facebook user will be perceived as more attractive and credible when they have more prosocial wall posts by physically attractive friends (Walther et al., 2008). Social media users who are highly motivated to manage their self-presentation must make decisions about how to react to and manage the comments, tags, and wall posts that others contribute. Therefore, the management of other-generated content will mediate the effects of such content on self-presentational content (see Figure 1). Adolescents who were interviewed by Barbovschi et al. (2018) indicated they did not have a lot of control over their self-presentation due to the contributions from others. Specifically, they identified that pictures from others in which they were tagged can contribute negatively to their online image. When others post content that social media users deem undesirable, decisions must be made about how to manage that content. These protective self-presentation behaviors can be repudiative, such as adding another post or photo, or subtractive, such as 87 2021, 9, 80-98 Hollenbaugh deleting or untagging the content (Rui & Stefanone, 2013). Rui and Stefanone (2013) found that Facebook users who reported they based more of their self-esteem on external validation, appearance, and/or competence were significantly more likely to engage in protective self-presentation when they received unwanted other-provided posts and photos. Untagging oneself from less desirable content is frequently used by more experienced Facebook users, especially those with stricter privacy settings, to manage impressions (Birnholtz et al., 2017). In a large-scale content analysis of log data from 50 million Facebook photo tags, Birnholtz et al. (2017) found that users were most likely to untag photos that were close-ups, did not include food or landscape, and had fewer likes or comments. People were also more likely to untag themselves in photos posted by others who were significantly older than 20 years (Birnholtz et al., 2017). which potential audience members actually read or viewed a post, they instead imagine that audience and craft their self-presentation with that audience in mind (Litt & Hargittai, 2016). For example, Marwick and boyd (2010) conducted a qualitative analysis of Twitter users, observing that although Twitter users typically recognize the amorphous nature of their audience, they are more likely to post to a bounded, identifiable audience. Similarly, a longitudinal diary study of over 100 Facebook, LinkedIn, and/or Twitter users found that about half of the time, participants imagined a specific audience when crafting their posts (Litt & Hargittai, 2016). Of those imagining a specific audience, 70% of the posts were intended for family and/or friends; however, participants reported fluctuating between various audience types from post-to-post (Litt & Hargittai, 2016). Litt and Hargittai (2016) argued that there are two sets of strategies social media users may employ to manage their audiences: audience-reaching and audience-limiting. Audience-reach- Research has examined the impacts that others’ likes and friend requests have on well-being and self-esteem (Burrow & Rainone, 2017; Meeus et al., 2019; Valkenberg et al., 2006), but less is known about the direct contributions of others on one’s presentation of self. Managing self-presentational content from others is a byproduct of the networked publics afforded by social media that is ripe for future research using a quantitative approach. Having reviewed research on affordances and other-generated content, it is now time to turn attention toward the complicated impacts that audience composition has on self-presentation. ing strategies, such as increased personal disclosures, are designed to draw in the target audience; on the other hand, audience-limiting involves excluding people who are not in the target audience, perhaps through managing who can view particular posts or removing someone from the network (Litt & Hargittai, 2016). The complex task for social media users is to manage potentially competing self-presentation goals in social networks with large, diverse audiences. The research reviewed hereafter suggests that the perceived audience will moderate the effects of self-presentation goal on content, such that under conditions of context collapse, selfpresentation content will be more carefully selected, selfcensored, and segmented (see Figure 1). Audiences and Context Collapse In social media, the audience – real or imagined – is often difficult to ascertain. Audiences are often imagined through consideration of the cues that are given off by the technology and the social context (boyd, 2007; Litt, 2012; Litt & Hargittai, 2016; Marwick & boyd, 2010). However, real and imagined audiences may be quite different. With social media tools and practices, such as sharing, tagging, and retweeting, the potential audience of a single post or tweet is amorphous (Marwick & boyd, 2010). That seems to especially be the case with Facebook, where the content originally shared to one’s audience might be made available to other networks when friends comment on, like, or share the content (Y. H. Choi & Bazarova, 2015). Regardless of the medium, because posters cannot be sure www.rcommunicationr.org Context collapse In imagining the audience within social media, widelyadopted social networking sites such as Facebook are often marked by context collapse, or “the flattening out of multiple distinct audiences in one’s social network, such that people from different contexts become part of a singular group of message recipients” (Vitak, 2012, p. 451). The invisible audience makes it impossible to fully assess the social context and background of the real audience, and given that content posted in social media is persistent, this presents challenges for social media users (boyd, 2011). As boyd (2011) explained, “In networked publics, contexts often collide such that the performer is unaware of audiences from different contexts, 88 Self-Presentation in Social Media magnifying the awkwardness and making adjustments impossible” (p. 51). Context collapse increases self-presentation concerns for social media users, and therefore moderates the relationship between self-presentation motivations and content (see Figure 1). Davis and Jurgenson (2014) proposed two different types of context collapse: context collusion and context collision. Context collusion involves intentionally drawing together varied audiences, which can have benefits to social capital (Davis & Jurgenson, 2014). Context collision, on the other hand, describes situations where the unintentional flattening of one’s social network results in “potentially chaotic results” (Davis & Jurgenson, 2014, p. 481). The potential for widely diverse audiences (e.g., friends, classmates, employers, professors, grandparents), coupled with the invisibility of that audience leads to confusion and often privacy concerns regarding verbal and visual posted content. context collapse, social media users may adopt a ‘lowest common denominator’ approach, where they post in ways that are socially appropriate for even the most socially distant person in their network (Hogan, 2010). There is qualitative support for the lowest common denominator effect (LCDE) in the location service Foursquare (e.g., Guha & Birnholtz, 2013; Saker, 2017). Study participants expressed concern with sharing late-night check-ins at bars for fear of the impression it would leave on parents (Guha & Birnholtz, 2013). Interview data suggested that Foursquare users often refrain from checking into locations that may reflect poorly on the image they wish to portray to important others in their network (Saker, 2017). Twitter users responding to Marwick and boyd’s (2010) survey reported refraining from topics that could be potentially controversial or too personal for some members of the imagined audience, as a form of self-censorship. Many social networking sites have developed tools to allow users to target particular posts to only certain others in their network, which can be used to facilitate audiencelimiting strategies (Litt & Hargittai, 2016). Although early research has shown that some people use these tools to manage the privacy of their content (e.g., Child et al., 2011; Ongun & Demirag, 2014), these tools may often be abandoned for ease of use. Segmenting the audience and reaching out to particular target audiences can help users manage context collapse (Litt & Hargittai, 2016). Stigmatized groups often experience intensified selfpresentation concerns in collapsed contexts. For example, LGBTQ Facebook users in Duguay’s (2016) qualitative study discussed their experiences with disclosing their sexual identity on the social medium. Some participants took advantage of the affordances of Facebook and, in an act of context collusion, intentionally broadcasted their disclosure of sexual identity. Others fell victim to context collisions when their sexual identities were involuntarily disclosed to segments of their networks whom they did not wish to tell (Duguay, 2016). Involuntary disclosures took the form of comments from others, page likes, group membership, and friends’ posts (Duguay, 2016). In most cases, and as would be expected, collapsed networks require more careful management of one’s self-presentation, which is often accomplished through self-censorship. According to boyd (2011), “The most controversial actors are those who hold power over the participant, such as parents, bosses, and teachers” (p. 44). Due to In contrast to the LCDE, Marder et al. (2016) have posed an alternate perspective, the strongest audience effect (SAE). Rather than accommodating the strictest audience, as with the LCDE, college students in Marder et al.’s (2016) study adjusted their disclosures according to the strict audience whom they valued the most. This newer model of audience effects in self-presentation has yet to be fully tested to see if it is a better predictor of behavior than the LCDE in social media. Utilizing decision-making tools such as the LCDE may jeopardize the success of self-presentation attempts for other segments of one’s network. Additionally, context collapse typically results in more cognitive efforts to manage varied self-presentation goals. For example, Facebook users with more diverse friends lists who received unwanted otherprovided posts and photos, such as being tagged in unattractive photos, tend to be more actively engaged in protective self-presentation (Rui & Stefanone, 2013). In other words, more heterogeneous networks may require Facebook users to un-tag, delete, or post alternative content in an effort to repair their images. Along with self-censorship via the LCDE or SAE, another option for managing context collapse is through segmentation of one’s network. Rather than broadcasting self-presentation content to one’s entire audience, social media users might participate in closed groups or use different social media accounts or apps to reach different types of audience. For example, politically active Norwegian teens in Storsul’s (2014) qualitative study explained that they 89 2021, 9, 80-98 Hollenbaugh joined Facebook groups dedicated to particular causes not only as a self-presentation strategy, but also to identify a relatively safe space to disclose their ideals. Within these more homogenous groups, they felt free to engage in political discussions, whereas they avoided commenting in those discussions open to their larger Facebook networks (Storsul, 2014). Another response to context collapse in social media with large, diverse audiences is to reconcile differences in selfpresentation and instead convey authenticity. Authenticity may be more objectively identified, such as through Twitter’s verification checkmark (Hearn, 2017), as well as through posted content. However, authenticity may further complicate the impression management process online because authenticity is also audience-specific (Marwick & boyd, 2010). What makes a person authentic to one audience may seem out of place or strange to another. Self-censorship may work for most people, but Twitter is offline. When there are inconsistencies between online and offline impressions, observers tend to judge their acquaintances quite harshly, more so than their friends (DeAndrea & Walther, 2011). DeAndrea and Walther (2011) conducted an experiment involving Facebook users’ identification and evaluation of inconsistencies in self-presentation for themselves, their acquaintances, and their friends. Participants rated their acquaintances’ inconsistencies as more intentionally misleading and were more likely to think they indicated more hypocrisy and untrustworthiness than the inconsistencies of their friends. However, even friends were judged more harshly than the participants’ own inconsistencies (DeAndrea & Walther, 2011). Given the prevailing research on audience effects, it is clear that the audience is central to the study of online selfpresentation and resulting content shared via social media. Research suggests that social media users with large networks have more opportunities to present themselves, but must balance the diversity of those networks. Impacted by the LCDE or perhaps the SAE, they may use self-censorship and segmentation while maintaining authenticity. As social media users consider the composition of their real or imagined audience, they can utilize a number of communication channels to present themselves. Although there are varied benefits to social media participation, social media users often balance their desire for self-presentation with concern for privacy amid collapsed networks (Vitak, 2012). People’s concern for privacy differs from one social media application to the next (Katz & Crocker, 2015; Quinn, 2014), with especially high concern for privacy in Facebook given the likelihood for context collapse (Y. H. Choi & Bazarova, 2015; Hollenbaugh, 2019). The research reviewed above suggests that audience composition, namely differences in context collapse, moderate the relationship between self-presentation motivations and content (see Figure 1). also frequently home to ‘micro-celebrities’ who are building and maintaining a large fanbase, often for commercial purposes (Marwick & boyd, 2010). Research suggests that these individuals balance professional content with some personal content (but not too personal) to appear authentically human on Twitter (Marwick & boyd, 2010). They attempt to appeal to the interests of the imagined audience to keep them connected. Often this balance is struck across posts, rather than within a post, such that tweets may appeal to different audiences in an effort to satisfy them collectively (Marwick & boyd, 2010). Qualitative research findings suggest that authors attempting to connect with their readership, as well as other authors, use personal self-disclosure to maintain authenticity in self-presentation (Laing, 2017). Social media users are authentic when they appear genuine in their social media posts, demonstrating that they are not trying too hard to secure a particular image of themselves. Posting very frequently, especially about mundane topics, could threaten one’s authentic self-presentation. In a recent study of Danish teenagers, too many check-ins, such as at a fitness center, were perceived as giving excessive effort. One study participant said, “…to me you can just feel the need for attention screaming out of all those updates” (Bertel, 2016, p. 169). This was the case for their participants especially when the check-ins were at mundane locations, such as school, work, and home (Bertel, 2016). Another threat to authenticity is when online impressions do not match what people know to be true about someone www.rcommunicationr.org Discussion The purpose of this review was to synthesize existing research on the impact of technological affordances and audiences on social media users’ self-presentation in hopes of sparking new directions for research and theoretical extensions. Traditional self-presentation theories can be built upon and expanded through considering many influencers, such 90 Self-Presentation in Social Media as audience composition. Based on this review of literature, several variables have emerged that should be considered in comprehensive and updated self-presentation theories. Social actors have near-limitless options for performing identity. Each social media channel brings with it particular social and technological affordances that may be salient in self-presentation processes. Therefore, the chosen channel(s) should be included in studying a model of self-presentation. The research reviewed above that affordances such as anonymity, persistence, and visibility likely moderate the relationship between one’s motivation to present the actual or ideal self and the self-presentation strategies used. Along with choices regarding their own content, social media users must also respond to content that others contribute through tagging and commenting on user content. Other-generated content, mediated by how social media users manage that content, will contribute to the content of one’s self-presenta- The modeled effects of these variables on self-presentation should be understood within the confines of particular limitations. First, most available research on online selfpresentation is qualitative, which provides a detailed description of the reality for many social media users, but is not meant to suggest generalized effects for the majority of people. Additional quantitative studies, informed by the qualitative research reviewed here, will provide more conclusive supporting or contradicting evidence. Second, although attempts were made to include all available research on the topic, this review may not be exhaustive. For example, other variables such as relational motives for using social media, gender identity, individual traits, culture, group membership, specific self-presentation content (verbal and visual), and feedback from audiences could be incorporated into a fuller conceptualization of social media self-presentation. Third, the reviewed research spans about 15 years, during tion. Finally, audience characteristics such as the size and diversity of the network, will moderate the relationship between presentation motives and content (see Figure 1). Traditional theories of self-presentation would benefit from an update for the social media age. The variables revealed in this literature review could be incorporated into the dramaturgical perspective (Goffman, 1959) and Schlenker’s (1985) theory of self-identification. For example, performances in social media spaces are often marked by higher amounts of persistence. Therefore, they may carry more weight than the fleeting performances of face-to-face impression managers. Additionally, context in social media spaces includes much more than the physical and temporal environment that was originally conceptualized; instead, audience composition and technological affordances play a central role in self-presentation in social media. which time social media has changed drastically. It is unclear whether research findings during social media’s early days would still hold true today. As a reminder, this provides more support for adopting the affordance framework when researching social media, but it is possible that the changing landscape of social media options and norms of use may cause one to challenge the findings from years ago. This review synthesizes research on self-presentation in social media, specifically channel- and audience-specific effects, to inform future research and theory in online selfpresentation. The features and uses of social media, as well as its ubiquitous presence in modern life, call for a reexamination and updating of traditional self-presentation theory. The aim of this paper is to contribute to the growing field by presenting a model that can be tested quantitatively to further uncover the complex processes of self-presentation in social media. Acknowledgement The author would like to thank the editorial staff and reviewers of Review of Communication Research for their thorough critique. Many thanks as well to colleagues in Writing Boot Camp who provided support and feedback throughout this project. Declaration of interest statement: This research received no funding from external sources. There is no conflict of interest, financial or otherwise, to report. 91 2021, 9, 80-98 Hollenbaugh References Anderson, M., & Jiang, J. (2018, November 28). 2. Teens, friendships, and online groups. In Pew Research Center: Internet & Technology. https://www.pewresearch.org/internet/2018/11/28/teens-friendships-and-online-groups/ Audrezet, A., de Kerviler, G., & Moulard, J. G. (2018). Authenticity under threat: When social media influencers need to go beyond self-presentation. Journal of Business Research. Advance online publication. https://doi.org/10.1016/j.busres.2018.07.008 Barbovschi, M., Balea, B., & Velicu, A. (2018). Peer-ing in the online mirror: Romanian adolescents’ disclosure and mutual validation in presentations of self on social media. Revista Româna de Sociologie, 29, 269-286. Bayer, J. B., Triêu, P., & Ellison, N. B. (2020). Social media elements, ecologies, and effects. Annual Review of Psychology, 71, 471-497. https://doi.org/10.1146/annurev-psych-010419-050944 Bertel, T. F. (2016). ‘Why would you want to know?’: The reluctant use of location sharing via check-ins on Facebook among Danish youth. Convergence: The International Journal of Research into New Media Technologies, 22(2), 162-176. https://doi.org/10/1177/1354856514543250 Birnholtz, J., Burke, M., & Steele, A. (2017). Untagging on social media: Who untags, what do they untag, and why? Computers in Human Behavior, 69, 166-173. https://doi.org/10.1016/j.chb.2016.12.008 Birnholtz, J., Fitzpatrick, C., Handel, M., & Brubaker, J. R. (2014). Identity, identification, and identifiability: The language of self-presentation on a location-based mobile dating app. Proceedings of ACM MobileHCI, Toronto, ON, Canada: Association for Computing Machinery. Blackwell, C., Birnholtz, J., & Abbott, C. (2014). Seeing and being seen: Co-situation and impression formation using Grindr, a location-aware gay dating app. New Media & Society, 17(7), 1117-1136. https://doi.org/10.1177/1461444814521595 Boyd, D. (2007) Why youth <3 social network sites: The role of networked publics in teenage social life. In D. Buckingham (ed.), Youth Identity and Digital Media (pp. 119-142). Cambridge MA: MIT Press. Boyd, D. (2011). Social network sites as networked publics: Affordances, dynamics, and implications. In Z. Papacharissi (Ed.), A networked self: Identity, community, and culture on social network sites (pp. 39-58). New York: Routledge. Brunswik, E. (1956). Perception and the representative design of psychological experiments. Berkeley, CA: University of California Press. Burrow, A. L., & Rainone, N. (2017). How many likes did I get?: Purpose moderates links between positive social media feedback and self-esteem. Journal of Experimental Social Psychology, 69, 232–236. https://doi.org/10.1016/j.jesp.2016.09.005 Carr, C. T., Hayes, R. A., & Sumner, E. M. (2018). Predicting a threshold of perceived Facebook post success via likes and reactions: A test of explanatory mechanisms. Communication Research Reports, 35(2), 141-151. https://doi.org/10.1080/08824096.2017.1409618 Chen, A. (2019). From attachment to addiction: The mediating role of need sat