Social Identities, Group Formation, and the Analysis of Online Communities

Home , Identity formation, Social facilitation, Social identity theory

1346

Chapter 5.4 Social Identities, Group Formation, and the Analysis of Online Communities

Jillianne R. Code Simon Fraser University, Canada

Nicholas E. Zaparyniuk Simon Fraser University, Canada

ABSTRACT collective identity and self-categorization associates social identity to online group formation. The Central to research in socialpsychologyisthemeans overall aim of this chapter is to explore how social in which communities form, attract new members, identity affects the formation and development of and develop over time. Research has found that online communities, how to analyze the develop- the relative anonymity of Internet communication ment of these communities, and the implications encourages self-expression and facilitates the for- such social networks have within education. mation of relationships based on shared values and beliefs. Self-expression in online social networks enables identity experimentation and development. INTRODUCTION As identities are fluid, situationally contingent, and are the perpetual subject and object of negotiation Central to research in socialpsychologyisthemeans within the individual, the presented and perceived in which communities form, attract new members, identity of the individual may not match reality. In and develop over time. The mechanisms in which this chapter, the authors consider the psychological communities grow depend on an individual’s challenges unique to understanding the dynamics ability to find and collaborate with others with of social identity formation and strategic interac- relevant knowledge, skills, and beliefs that meet a tion in online social networks. The psychological particular need. While these mechanisms of social development of social identities in online social collaboration are not unlike traditional face-to-face network interaction is discussed, highlighting how interactions (Tyler, 2002), there are some important differences in the way in which group members interact in online environments. Relative anonym- DOI: 10.4018/978-1-60566-208-4.ch007

Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Social Identities, Group Formation, and the Analysis of Online Communities

ity,selectiveself-disclosure,physicalappearance, ferent social groups, results in a novel dynamic to and the ease in finding ‘familiar others’ through human community formation and evolution. search, embedded traits, and predefined groups, In this chapter, we consider the psychological are some of the important differences between challenges unique to understanding the dynam- Internet communication and face-to-face interac- ics of social identity formation and strategic tions(Bargh&McKenna,2004;McKenna,Green, interaction in online social networks. We start & Gleason, 2002; Walther, 2007). Research into with a brief overview of aspects within social Internet social interaction has led to an increased psychology that are pertinent to a discussion on understanding of face-to-face communications social identity formation in online social net- andbringsintofocustheimplicitassumptionsand works. Specifically, we introduce Social Identity biasesthatexistintraditionalcommunication(Lea Theory as a perspective in which to frame our & Spears, 1995; Tyler, 2002). Assumptions that current understanding of online social network mediate face-to-face interactions such as physical formation. Next, the psychological development proximityandnon-verbalcues,assumednecessary of social (virtual) identities (Jenkins, 2004) are to communicate and relate, do not exist in most explored in online social networks using the Internet communications. However, given these conceptualization of self-presentation (Goffman, limitations, online social communities continue 1959/1997).Adiscussionofcollectiveidentityand to thrive and grow. The evolution of online com- self-categorization follows and relates how social munities confronts current views of how social identity contributes to onlinegroup formation and and psychological dynamics contribute to human evolution. Further, to illustrate how to evaluate relationships, communication, and community the effectiveness of online social networks, we formation. review several studies on online social networks Research supports the idea that the relative using ethnographic methodologies, visualization anonymityofInternetcommunicationencourages techniques, and social network analysis (SNA). self-expression and facilitates the formation of Finally,wepresentpracticalteachingandlearning relationships outside of what is considered ‘nor- strategies educators can use to facilitate the use of mal’socially mediated communication (Wallace, socialsoftwareforonlinesocialnetworkformation 1999). The complex origins of shared values and withineducationalenvironments.Theoverallaim beliefs(Bargh&McKenna,2004),self-expression of this chapter is to explore how social identity through identity experimentation (Ruitenberg, affects the formation and development of online 2003), and relative anonymous interaction (i.e. communities, to present some methodologies for strangers on the train effect; Derlega & Chaikin, evaluating the effectiveness of group formation, 1977; Rubin, 1975) challenge ideas of an ‘indi- and to explore the implications of online social vidual’ identity in relationship formation (Lea & networks within education. Spears, 1995). As individual identities are mal- leable, adaptable, and the perpetual subject and objectofnegotiationwithineachcontext(Jenkins, SOCIAL IDENTITY AND 2004), the notion of identity requires an incessant THE INTERNET comparison between the individual, the context in which they are interacting, their intentionality All human identities are social identities (Jen- in the context of that interaction, and their ‘true’ kins, 2004). Social identity concerns how we (nominal) identity. The irregular nature in which identify our similarities and differences to other individuals present arbitrary identities in various known groups of individuals. Social identify is contexts, with multiple intentions, and within dif- an ongoing interplay between how we identify

1347 Social Identities, Group Formation, and the Analysis of Online Communities

ourselves and how others identify us. To identify tional significance attached to it” (Tajfel, 1981, p. with any given group of people, whether it is an 255). As individuals belong to a variety of social ethnic group or an online organization, we look groups, their overall self-concept is composed of for similarities between the group members and multiple socialidentities(Ashforth&Mael,1989; ourselves. While similarities initially attract an Hogg, Terry, & White, 1995; James, 1891/1950; individualtoagroup,thisinitialattractionenables Jenkins, 2004). These multiple social identities an individual to recognize their individual differ- enable an individual to adopt various roles and ences. This comparative process is identified in adapt to a variety of social contexts. The contexts Social Identity Theory as “internal and external in which a social identity exists, supports the plu- moments of dialectic identification” (emphasis ralistic nature of the Self. As social groups exist at in original; Jenkins, 2000, p. 7). These internal multiple levels, i.e. societal, cultural, industrial, or- comparisons are how individuals distinguish ganizational, functional, and professional (Korte, themselves from others in both the similarities 2007), individual’s social identities are facilitated they share and the differences they recognize. through communication withinand amongstthese Alternatively, external comparisons involve how levels. For example, a professor can identify him others identify individuals, in the similarities and or herself as teacher, parent, friend, advocate, and differences they see between that individual, administrator based on social context. The use themselves, and a particular group. As internal of Internet-based communication technologies, and external comparisons determine the active such as Internet messaging (IM), chat, and social and socialized aspects of a person, they enable networks, provide an extension of social contexts the differentiation of the I and me which make in which individuals can interact. The various up an integrated Self (James, 1891/1950). Given social context and relationships developed using that social identification involves the interplay such social technologies, facilitates the develop- of internal and external dialectic processes, the ment and recognition of an individual’s social Internetfurtherenablesindividualstodevelopand identification. express multiple social identities and experiment with new virtual identities. As individuals’social Social Identification identities evolve from within social groups, they also facilitate the alignment or differentiation of Social identification is an emergent product of an individual from the group. This alignment or internal-external dialectic processes (Jenkins, differentiation reaffirms an individuals’ social 2004).Emphasizingadistinctionbetweeninternal identity. and external dialectical processes (Barth, 1969) Social identity is a central construct in under- allows a “wider distinction to be drawn between standing intergroup relations and is a key element nominal identity and the virtual identity: between in linking an individual to his or her social group the name and the experience of an identity” (em- (Tajfel, 1974, 1981). According to Tajfel (Tajfel, phasis added; Jenkins, 2004, p. 22). A nominal 1974, 1981), the foundation of the Social Iden- identity is the label an individual identifies his tity Theory of group membership and behavior or her Self with, and a virtual identity is the ex- recognizes that grouping (social categorization) perience of that nominal identity. In other words, influences people’s perception of others and one’s your nominal identity is who you believe you Self.Socialidentityis“thatpartoftheindividual’s are (internal dialectic), and your virtual identity self-concept [or self-identity] which derives from is the experience of being that person (external his or her knowledge of membership to a social dialectic). In addition, nominal identification var- group(orgroups)togetherwiththe value and emo- ies from context to context and can be associated

1348 Social Identities, Group Formation, and the Analysis of Online Communities

with numerous virtual identities (Jenkins, 2004). explore identities that they could not explore in For example, one may identify him or herself as a their ‘regular’ everyday lives or in their youth student(internalidentification,nominalidentity), (Archer, 1989). but his or her identity and experience as a student Through adolescence, the uncertainty of iden- is quite different from high school to university tity dominates an individual’s development and (externalidentification,virtualidentity).Similarly, definition of who they feel they are. As identity is the same student may consider him or herself a aprocessof‘becoming’,identityexperimentation quiet,shypersoninface-to-facemeetings(nominal becomes a means of self-exploration. A recent identity), however, when online they present them- study by Valkenburg, Schouten, & Peter (2005) self as having an outgoing, animated personality investigated identity experiments by adolescents (virtualidentity).Theexperienceofanindividual’s and whether pre- and early adolescents engage identity, as perceived through thought and action, more often in Internet-based identity experimenta- isinfluencedthroughtheinterplayoftheindividual tionandself-presentationalstrategiesthanmiddle and social others. The evolution and development and late adolescents. Of the 600 adolescents sur- of internal and external dialectic processes occurs veyed (ages 9-18, M = 13.37, SD = 1.98), 82% through social identity experimentation. indicatedthattheyusedchatorInternetMessaging (IM) on a regular basis. Of those who used these Social Identity Experimentation technologies, 50% of them reported that they willingly experimented with their identity. Using Experimenting with social identities is an impor- such self-presentation strategies as presenting his tantpartoflifespandevelopment(Wallace,1999). or herself as older, more ‘macho’, more ‘beauti- As individuals develop, particularly through ful’, more ‘flirtatious’, as the opposite gender, as adolescence, they begin to question their place a real-life acquaintance, or as a ‘fantasy’ person, in society; leading them to question their identity these adolescents were actively engaged in con- andpersonalvalues(Erikson,1963,1980).Within scious identity experimentation. The majority of today’sfastpacedenvironment,wherelifestyleand the adolescents surveyed in the study (49.8%) career options are abundant and change quickly, presented themselves as older. Further findings many individuals return repeatedly to question reveal that relative to age differences in the group their values, beliefs and life goals (Archer, 1989; surveyed, there was a strong influence of age on Wallace, 1999). In particular, the Internet plays Internet-based identity experimentation (b = -.50, an important role in social identity formation and p<.001).Meaning,thatyoungeradolescencesare development as it allows individuals to explore significantly more likely than older adolescences their values and beliefs within environments that to experiment with their socialidentities,andmore they perceive to be safe. The anonymity of online frequently use their social identities to facilitate interactions facilitates the perception of safety of social interaction. Valkenburg et al. also report anindividual’snominalidentity,allowingusersto that introverts engage in identity experiments as experiment with multiple virtual identities. As the social compensation more often than extraverts Internet expands opportunities for social identity do. The results indicate that introverts were more experimentation, through online chat, massively likely to experiment with their social identity multiplayer online games (MMOG), 3D online as a means to explore social communication virtual worlds, and social networks, individuals they lack in the face-to-face world. The study readily test and experiment with multiple identi- concludes that older teenagers used the Internet ties. The ambiguity of the Internet in one’s life most often to communicate with their existing course, is an enabling factor for individuals to personal network, whereas younger adolescents

1349 Social Identities, Group Formation, and the Analysis of Online Communities

are more likely to use the Internet more frequently as cited in Abrams & Hogg, 2001). According to to communicate with strangers. The results of the Goffman (1959), individuals present an impres- Valkenburg et al. study validate some existing as- sion by performingand observers are the audience sumptions about why adolescents use the Internet that judges the effectiveness or the believability as a predominate medium of communication, and of the performance. Goffman describes the per- provides additional evidence of Internet identity former-audience dialectic as one concerning the experimentation as a means for uninhibited self- maintenance of the impression that the performer exploration is ‘living up’ to the standards by which their ac- Motives for identity experimentation are var- tions are judged. Whether or not these actions ied and diverse. Self-exploration (i.e. to explore are true of the individual’s identity, remain the how others react), social compensation (i.e. to subject of the audience’s judgment. Studies re- overcome shyness), and social facilitation (i.e. to veal that in virtual settings, such as online social facilitaterelationship-formation)areafewmotives networks, inaccuracy of interpretation resulting foridentityexperimentation.Identityexperiments from individual presentation is a major challenge such as the one explored in the Valkenburg et al. with Internet communication (Donath, 1999; study, demonstrate the reciprocal nature of Self Lynch, 2005; Valkenburg et al., 2005; Walther, and social group interaction in the formation of 2007). As previously discussed in the study by identity. For an individual to develop a social Valkenburg et al. (2005) and in the literature on identity, what that individual thinks of him or identity experimentation (e.g. Wallace, 1999), the herself is significant, but no less significant than Internet presents many opportunities and motives what others think about him or her. To return to the foridentityexperimentation.Asmostparticipants internal-external dialectic discussed previously, in online social networks are likely to be actively what the Valkenburg study demonstrates is that experimenting with different social identities, it is not enough to assert a social identity; others ‘audience’members need to be aware that people must also validate that social identity through its may be presenting an identity that may only be receptionandrecognition.Self-presentationthen, a small part of the ‘performers’ nominal identity is an assertion of a social identity. (e.g. Walther, 2007). This awareness brings literal meaning to Shakespeare’s assertion that “All the Self-Presentation world’s a stage, and all men and women play- ers” (As You Like It cited in Haney, Banks, & Self-presentation is an individual’s projection of Zimbardo, 1973). Self and identity in the social world (Valkenburg et al., 2005). In a traditional face-to-face set- ting, the reality that the individual is concerned GROUP FORMATION with is generally unperceivable. The individual observes the situation and acts according to their Group membership is crucial to the internal- perceptions; even if their perceptions are inac- external dialectic negotiation that is identity curate. “Paradoxically, the more the individual (Amiot,delaSablionniere,Terry,&Smith,2007). is concerned with the reality that is not available Self-categorization theory (Turner, 1985, 1987) to perception, the more must he concentrate his suggests that identification with any group is attention on appearances” (Goffman, 1959/1997, based on the extent to which individuals can en- p. 21). People can change their persona to reflect hance their social identity through categorizing the social audience and can have as many social themselves as group members (Chattopadhyay, ‘selves’ as there are situations (William James George, & Lawrence, 2004). This theory suggests

1350 Social Identities, Group Formation, and the Analysis of Online Communities

that individuals must associate themselves and deems as deviant behavior or for ideas that are others with particular social categories to derive contradictory to the norms of the group (Marques, social identities (Turner, 1985). Abrams, Paez, & Hogg, 2001). Ultimately, self- categorization depersonalizes perception and Self-Categorization conduct such that individual members are not ‘processed’ as complex multidimensional whole Social identity involves a process of self-catego- persons, but rather as embodiments of the group rization. Categorization as a cognitive function prototype (Birchmeier, Joinson, & Dietz-Uhler, enables individuals to perceive the world as 2005; Chattopadhyay et al., 2004; Haney et al., structured and predictable. Categorization is one 1973). Research has found that social and group of the most basic and essential of all cognitive identities are generally more powerful than in- processes that helps one focus on contextually dividual identities, and there is a tendency for relevant and meaningful aspects of the world; individuals to go along with the group in which highlighting important distinctions and de-em- they identify (Hogg & McGarty, 1990; Korte, phasizing unimportant ones (Hogg, 2001). For 2007; Tyler, 2002). The sense of group identity example, a student may categorize himself or and the degree of personal identifiability to other herself as a football player, or other students may groupmembersareconditionsknowntoinfluence categorize that student as a football player. Given this power relationship (Taylor & MacDonald, this‘footballplayer’categorization,students(and 2002). There is also a tendency of the individual even teachers) make certain assumptions about to downplay personal attributes in favor of the how that student is likely to behave, with whom group prototype or collective identity. he or she associates, and even his or her ability for academic achievement. Categorization of Collective Identification Self, relative to group membership, emphasizes perceived similarities among group members and Collective identification is a representation of the characteristics that best define the group in how people are similar to each other based on the that particular context (Hogg, Cooper-Shaw, & psychological connection between Self and social Holzworth,1993).Self-categorizationaccentuates group (Abrams & Hogg, 2001; Jenkins, 2004).As attitudinal,emotional,andbehavioralsimilarityto discussed previously, socialidentityisapartofthe a group prototype (Hogg & Hains, 1996). Agroup Self that one identifies with a particular category prototype involves the salient characteristics that or group. Put another way, social identity is “the define a typical member of that group. As a pro- perceptionofselfintermsofstereotypicalingroup totype is shared amongst group members, it also (sic)attributes”(emphasisadded;Abrams&Hogg, identifies group norms and stereotypes (Hogg et 2001, p. 433). The in-group is simply the group al., 1993). For example, the prototypical football in which one identifies, conversely, the out-group player is an individual who has superior physi- are those individuals who are not exclusive mem- cal abilities, is disruptive in class, and does not bers. Collective identification, thus, results in a obtain high grades in academic subjects. Further strong association between an individual and the to our prototypical football player, if this student group in which they are member. The individual deviates from what is stereotypical or ‘normal’ then assumes the collective identity. Barnum and for this group, such as achieving high grades in Markovsky (2007) hypothesized that in-group their academic subjects, they may be subjected memberswouldbemoreinfluentialthanout-group to ridicule from their peers. A group may ostra- members on the collective. For example, using cize a fringe member based on what the group two theoretical approaches based on disagree-

1351 Social Identities, Group Formation, and the Analysis of Online Communities

ments with in-group members (self-categorization Turner, 1987) in which group membership (col- theory) and the performance expectations (status lective identification) is facilitated, is important characteristics theory) of the in-group members, for understanding the dynamics of individuals Barnum and Markovsky (2005) observed that acting as a collective unit or group. The mixed group membership affected social influence and results of the Postmes et al. study reveal that both that in-group members influenced their subjects individuals and groups are in a constant state of more than out-group members. The results of this social flux. The dynamic nature and negotiation studysupporttheargumentthatinthedevelopment of these groups online makes their structure and of a social identity, the new group or collective evolution fluid and uncertain. Tools, techniques, identity tends to depersonalize the individual in and technologies for analyzing social networks, favor of becoming a group member. will enable further our understanding in social identity development and group formation, and Depersonalization aidindeterminingthemeasurableimpactofsocial network tools in education. Depersonalization causes people to conform to group prototypes and behave according to group norms. Similar to deindividuation of iden- ANALYSING ONLINE tity (Festinger, Pepitone, & Newcomb, 1952; SOCIAL NETWORKS Zimbardo, 1969), depersonalization gives an individual a sense of anonymity, in which he or Social network formation is a complex process in she submits themself to the collective identity. whichindividualssimultaneouslyattempttosatisfy Postmes,SpearsandLea(2002)hypothesizedthat goalsundermultiple,oftenconflicting,constraints depersonalization would increase the tendency (Kossinets&Watts,2006).Socialnetworkanalysis for intergroup differentiation in attitudes and (SNA) involves theorizing, model building and stereotypes specifically with computer mediated empirical research focused on uncovering thepat- communications (CMC). Based upon previous terns of links among network members (Freeman, research(Postmes,Spears,&Lea,1998),Postmes 2000). Social network analysis conceives social et al. state that communication via CMC would structure as a social network. A set of social ac- potentiallyincreasedifferentiationbetweengroups tors and a set of relational ties connecting pairs on dimensions of bias, stereotyping, and attitude of these actors (Wellman, 2000) forms the social divergence. In addition, they postulate that CMC network. Social network structures are analyzed shifts intergroup interactions from interpersonal usingmeasuressuchasdensity,centrality,prestige, (“me” and “you”) to intergroup (“us” vs. “them”) mutuality, and role. Demographic data, such as ultimately depersonalizing interactions and stimu- age, gender, and ethnicity, and information about latingatendencyfordifferentiationbetweensocial ‘user’attitudes and beliefs are collected to gain an categories (Postmes et al., 2002). Postmes et al. understanding of the ethnographic characteristics (2002) could notattributedifferences in their find- of group members. Methods used in SNAinclude ings between the groups studied, rather, that CMC graph theoretic, algebraic, and statistical models likelyaccentuateddifferencesthatalreadyexisted. (Wellman, 2000). Due to the focus and length Postmes et al. claim that the results of their study constraints of this chapter, the specifics of SNA were heightened because of the online context methodologies and analysis are not explored in- despite the group differences that already existed. depth. Instead, we focus on the analysis of online The transition from the personal (nominal) to the socialnetworksusingexamplesfromtheliterature social(virtual)identity(asoriginallypostulatedby that consider linking individuals with community

1352 Social Identities, Group Formation, and the Analysis of Online Communities

growth, ethnography, social discourse, and data social enterprises. As online social networks visualization. offer commercial advertising space to a captive audience and is a rapidly evolving environment Examining Links for social research, understanding how, why, and under what conditions these groups thrive is of Examininglinksbetweengroupmembersenables paramount importance. For example, a recent researchers to understand how individuals influ- study by Backstrom, Huttenlocher, Kleinberg, & ence, relate, and interact in social networks. Kos- Lan (2006) explored three questions in regards sinets & Watts (2006), in an analysis of a dynamic to online social network growth and evolution. social network of more than 45,000 students, First, they considered membership and the struc- faculty, and staff at a large university, found that tural features that influence whether a given indi- networks evolve as a result of effects arising be- vidual will join a particular group. Second, they tweenthenetworktopologyandtheorganizational examined how structural features that influence structure the network embodies. Of particular a given group and whether that group will grow interest is that network characteristics (measures) significantly over time. Finally, Backstrom et al. appear to approach an equilibrium state, whereas explored aspects of group change and how group individual properties such as linking and bridg- foci or topics change over time and whether this ing are considerably more complex and are more dynamic affects underlying group membership. appropriately analyzed using ethnographic tech- Backstrom et al. found that the formation of niques (as discussed in a later section). Linking groupsandthedeterminingfactorsofmembership and bridging of individuals-to-individuals and significantly relate to the internal connectedness groups-to-groups facilitate connections outside of an individual’s friends. Meaning, individuals of an individual’s circle of acquaintances and whosefriendsareinacommunityaresignificantly promotes a diffusion of information and growth more likely to join that community. In a similar of existing and new communities (Kossinets & way to bridging, as discussed in the last section, Watts, 2006). The rapid and dynamic nature of information diffusion is similar to membership linking and bridging in the growth and develop- diffusion in that the more links or bridges one ment of social networks within a relatively stable obtains affects the development of the social infrastructure is recognized in the rapid growth network and expedites its growth. Backstrom et of websites such as Facebook (2008), MySpace al. also examined the flow of information within (2008),SecondLife(LindenResearchInc.,2008), groups; specifically they questioned that “given a and Bebo (2008). Within these web communities, set of overlapping communities, do topics tend to social network connections are far more complex followpeople,ordopeopletendtofollowtopics?” thanthetechnologicalinfrastructureinwhichthey (Backstrom et al., 2006, p. 8). The results to this are situated. Understanding how and why bridges final query were inconclusive indicating that less occuriscentralinunderstandingthecircumstances technicalapproachestounderstandingcommunity surrounding the formation and growth of online formation and growth, such as the methods used communities. in ethnographic research, would likely provide clearer answers as to the complex dynamics that Research on Community Growth take place in online social networks. in Online Social Networks

Community growth in online social networks is of great importance to both commercial and

1353 Social Identities, Group Formation, and the Analysis of Online Communities

Ethnography and Social Discourse the homophily principle that similarity breeds connections to “people like us” (McPherson, Ethnography is a method of research primarily Smith-Lovin, & Cook, 2001). Homophily, as concerned with the description of natural hu- demonstrated by the boyd and Heer study, serves man communities (Munroe, 2000). Ethnography as a limiter in individuals’social world. As social enables the interpretation of the flow of social networks consist of people who know each other discourse (Gertz, 1973/2000). In the study of offline and who are similar in sociodemographic, online social networks, ethnography is particu- behavioral,andintrapersonalcharacteristics,these larly useful in studying online groups as unique networks are less dynamic and are more often a cultural communities. The methodologies and digital representation of most face-to-face social perspectives of ethnography, aids in establishing groups. As individuals interact with others similar new questions for research in social networks and to themselves (Baym & Zhang, 2004; Jones & complements existing quantitative methodolo- Madden, 2002) and attempt to avoid conflicting gies. A recent study by boyd & Heer (2006) used relationships (Bargh & McKenna, 2004; Gross, ethnographic techniques to study the dynamics of 2004), homophily is limiting because it prolifer- the popular international social networking site ates the divides in our personal environments and Friendster (2008). The ethnographic components limits exposure to people and networks different consisted of a 9-month participant observation, from our own. Finally, boyd and Heer describe including interviews, qualitative surveys, and a phenomenon they call “negotiating unknown focusgroups.BoydandHeer’sparticularresearch audiences” (boyd & Heer, 2006, p. 4); meaning questions involved examining how context is as users generate online contexts to serve the created and interpreted in digital environments, needs of a particular group, the individuals in how conversations are initiated online, what are those groups come together already associated the goals of digital conversations, and how are the group. Informed by their ethnographic inves- they maintained. Exploring the possibilities and tigation, boyd and Heer used data visualization to consequences of replicabilitiy, searchability, and provide a macroscopic view of many of the most persistence, boyd and Heer’s ethnographic study common behaviors they observed, such as brows- revealedseveralinterestingfindings.First,inorder ing photos, exploring profiles, and searching for to derive contextual cues in lieu of the physical common interests. As qualitative and quantitative cues present during face-to-face interactions, analysis of social networks provides insight into members of the social network interpreted what the interactions of individuals within the group, boyd and Heer describe as “artifacts of digital data visualization enables a macroscopic view of performance.” The ‘artifacts’ they describe are the social networks in question. traces of interaction history (Wexelblat & Maes, 1999), such as previous discussion postings and Data Visualization images. These artifacts served existing and new network members who use these virtual cues to Visualization aids in the presentation of abstract interpretandbuildasocialprofileoftheindividual data. Data visualization enables a visual means who left them. Second, as individuals invited ex- to confirm observations made at a local (in this isting friends to their social network, the groups case individual) level, as in the boyd and Heer grew and quickly became homogenous. Although study (2006), but also provides an alternative boyd and Heer infer that homogeneity is due to perspective on the patterns the data presents, as the limiting nature of the website itself, the emer- in a recent study by Golbeck (2007). In the boyd gence of a homogenous social group exemplifies and Heer (2006) study, the visualization served

1354 Social Identities, Group Formation, and the Analysis of Online Communities

particularlyusefulinconfirmingtheethnographic STRATEGIES FOR USING SOCIAL observations concerning the presence and com- SOFTWARE IN EDUCATION position of network clusters which allowed the researchers to develop additional narratives. An Social networks already exist in education. exampleofthevisualizationpresentedintheboyd Sports teams, social clubs, cheer squads, and and Heer (2006) and Heer and boyd (2008) and social cliques are a few examples that are rec- represents a single user profile, and demonstrates ognizable in any educational institution. As the theinterconnectednessbetweenhisorherexample context in which social identity supports the profile and their ‘friends.’ pluralistic nature of the Self, the learning envi- A recent study by Golbeck (2007) used visu- ronment is a particularly appropriate place for alization techniques to analyze social network students to explore their nominal identities and membership and relationship dynamics. The experiment with new social (virtual) identities. visualization Golbeck uses in the study of social For educators to capitalize and facilitate identity network growth of a sample of social networks experimentation within online social networks, show a steady linear growth rate. As awareness such as Facebook, MySpace, Second Life, and of the networks existence grew, mostly through Bebo, they need to facilitate social interaction advertising, the membership among the selected in all learning contexts. We have identified four networksgrewrapidlyfrom1000memberstomore important areas of research that support identity than 10 million. In analyzing the rate of relation- experimentation and promote the use of online ship growth relative to membership growth, the social networks in education. Research important ‘spacing’ of relationships increased significantly in the use of online social networks in education overtime.‘Spacing’suggests thatsocialnetworks involves investigating ways in which educators become more densely connected as they grow can preserve relative anonymity, enable identity larger, which was also observed in the Backstrom experimentation, manipulate self-categorization, et al. (2006) study discussed earlier. Golbeck used and measuring the effectiveness of online social visualization to illustrate this spacing effect. networks in education. The analysis of online social networks from the macro level provides specific information on Preserve Relative Anonymity community growth, social discourse, and general group dynamics. The addition of this macro-level To establish an equality of participation, relative information to the research on social identity pro- anonymity should be preserved (Taylor & Mac- vides further information on the ongoing interplay Donald,2002).Inorderforonlinesocialnetworks between the groups in which we identify ourselves to be successful in an educational context, ano- and how others respond to this dynamic. The rela- nymity needs to be maintained until social links tive strength of collective associations within the are established. If anonymity is not maintained, group, as evidenced through density, linking, and then the social network is likely to fail given bridging, associates individual activity with collec- that relatively few will participate, and if they tive (group) level properties.Through this process, do participate, they will ‘self-present’ in such a further detail of group prototypes and the salient way as to make the environment seem false. For characteristics that define a typical member of that example, in traditional or formal learning set- group can be identified, and additional ‘narratives’ tings, a student will often demonstrate behaviours can be developed that further understanding of so- expected by the teacher as opposed to behaving cial identity and group dynamics at the micro (i.e. as they actually feel. In an online environment, individual) and macro (i.e. community) levels. students should feel that they have the freedom to

1355 Social Identities, Group Formation, and the Analysis of Online Communities

express and interact in ways that are not reflective categorization, students are able to recognize how of outside social influence. The use of anonymity their participation within a socialgroupaffectsthe hasimplicationswithinformallearningstructures social network and remain accountable for their where anonymity is not often preserved. Further actions. The use of roles within group interactions research in this area is necessary to delineate enables individuals to understand the difference teaching and learning strategies for use within of their nominal identity and virtual identity in formal learning contexts. social contexts. This realization can only help students become more cognizant of the influence Enable Identity Experimentation of social roles in identity formation.

Encouraging identity experimentation facilitates Measuring Social Network the development of social networks that con- Effectiveness tinually evolve and change with each different educational context. Identity experimentation, Ameasurable means for observing changes in the keeping a modicum of anonymity, enables a various properties of a social network, such as student to present various social identities to his bridging,linking,andspacingwithinthenetwork, or her peers and the instructor. In other words, is important in helping educators determine the identity experimentation through various modes level of interactivity in the network overall. For ofself-presentationencouragesexpressionsofself example, a network with low interactivity would that are accepted or rejected by members of the have few bridges, few links, and very large spaces in-group.Identityexperimentationisimportantto betweenindividualsinthenetwork.Alternatively, education, as it is something that all individuals anactivenetworkisonewithhighinteractivity,has ‘do’and is uniquely possible within online social bridges, several links, and has less space between networking. Incorporating strategies within the individuals in the network. From an educational educational context that encourage such experi- perspective,anactiveclassroomnetwork,whether mentation in a safe and equitable way will foster face-to-face or online, that has many bridges and tolerance and understanding of other differing many links is a more productive learning environ- pointsofview.Identityexperimentationpromotes ment. It is the ability of the instructor to facilitate social interaction as the individual’s identity is in these links that will provide an effective social constant negotiation between the individual and learning environment. social group.

Manipulate Self-Categorization CONCLUSION

Building on the first two factors results in a ma- Experimenting with social identities is an impor- nipulation of self-categorization. Manipulating tantpartoflifespandevelopment(Wallace,1999). self-categorizationraisesindividualandcollective As individuals develop and change, they question awarenessof the variouseffectsof collective iden- their place in society leading them to question tification. In other words, teaching and learning their identity and personal values (Erikson, 1963, strategies that enable self-experimentation will 1980). People often change their persona (James, also influence how a student develops aware- 1891/1950) and have as many social ‘selves’ as ness of their impact on others in their peer group situations (Abrams & Hogg, 2001) and social and illustrates how they are accountable for groups. Categorization of the Self relative to their actions. Through the manipulation of self- a group accentuates the perceived similarity

1356 Social Identities, Group Formation, and the Analysis of Online Communities

between individual group members and one’s facilitate participation, experimentation, and re- representation of the features that best define the search of socialidentity,willultimatelycontribute group in a particular context (Hogg et al., 1993). to student’s insight into the dynamics of learning Arepresentation of how people are similar to each and development as a social process. other, is based on the psychological connection between the self and social group (Abrams & Hogg, 2001; Jenkins, 2004). Collective identity REFERENCES aids in the development of social identity, but tends to depersonalize the individual in favor of Abrams, D., & Hogg, M. A. (2001). Collective becoming a group member. Although deperson- identity: Group membership and self-conception. alization facilitates a transition from a personal In M. A. Hogg & S. Tinsdale (Eds.), Blackwell (nominal) to social (virtual) identity, where group handbook of social psychology: Group processes membership (collective identification) becomes (pp. 425-460). Malden, MA: Blackwell. increasingly important, social network formation Amiot, C., de la Sablionniere, R., Terry, D., & is a complex process in which many individuals Smith, J. (2007). Integration of social identities in simultaneously attempt to satisfy their goals un- theself:Towardacognitivedevelopmentalmodel. der multiple, conflicting constraints (Kossinets Personality and Social Psychology Review, 11, & Watts, 2006). 364–368. doi:10.1177/1088868307304091 The use of Internet-based communication technologies facilitates the development of social Archer, S. L. (1989). The status of identity: Re- groups and social identification. Social identity is flections on the need for intervention. Journal of central in understanding intergroup relations and Adolescence, 12, 345–359. doi:10.1016/0140- is a key element linking individuals to their social 1971(89)90059-6 group (Tajfel, 1974, 1981). As social groups exist Ashforth, B. R., & Mael, F. (1989). Social at multiple levels, social identity development is identity theory and the organization. Acad- facilitated through communication within and emy of Management Review, 14(1), 20–39. amongst these levels. Social identification, as an doi:10.2307/258189 emergent product of internal-external dialectic processes (Jenkins, 2004), enables an individual Backstrom, L., Huttenlocher, D., Kleinberg, J., & to experiment with different virtual identities and Lan, X. (2006). Group formation in large social explore what it is like to experience those identi- networks: Membership, growth, and evolution. ties in the social world. Paper presented at the KDD ‘06, Philadelphia, Motives for identity experimentation, such as USA. self-exploration (i.e. to explore how others react), Bargh, J. A., & McKenna, K. Y. A. (2004). social compensation (i.e. to overcome shyness), The Internet and social life. Annual Review of and social facilitation (i.e. to facilitate relationship- Psychology, 55, 573–590. doi:10.1146/annurev. formation) are all important factors in social psych.55.090902.141922 development. As schools are inherently social institutions, the factors that contribute to healthy Barnum, C., & Markovsky, B. (2007). Group mem- social development are of paramount importance bership and social influence [Electronic Version] for educators to consider for students growth. The [from http://www.uiowa.edu/~grpproc]. Current strategies for social identity experimentation in Research in Social Psychology, 13, 1–38. classrooms, allows students to become active interpreters of social interaction. Educators that

1357 Social Identities, Group Formation, and the Analysis of Online Communities

Barth, F. (1969). Ethnic groups and boundaries: Facebook, Inc. (2008). About Facebook.Retrieved The social organization of cultural difference. April 21, 2008, from http://www.facebook.com/ Oslo, Norway: Universitetsforlaget. about.php Baym, N. K., & Zhang, Y. B. (2004). Social Festinger,L.,Pepitone,A.,&Newcomb,T.(1952). interactions across media: Interpersonal com- Someconsequencesofde-individuationinagroup. munication on the Internet, telephone and face- Journal of Abnormal and Social Psychology, 47, to-face. New Media & Society, 6(3), 299–318. 382–389. doi:10.1037/h0057906 doi:10.1177/1461444804041438 Freeman, L. C. (2000). Social network analysis: Bebo, Inc. (2008). About Bebo. Retrieved April Definition and history. In A. E. Kadzin (Ed.), Ency- 21, 2008, from http://www.bebo.com/StaticPage. clopedia of Psychology, 7, 350-351. Washington, jsp?StaticPageId=2517103831 DC: American Psychological Association. Birchmeier, Z., Joinson, A. M., & Dietz-Uhler, Friendster, Inc. (2008). About Friendster. Re- B. (2005). Storming and forming a norma- trieved April 21, 2008, from http://www.friendster. tive response to a deception revealed online. com/info/index.php Social Science Computer Review, 23, 108. Gertz, C. (1973/2000). The interpretation of cul- doi:10.1177/0894439304271542 tures. New York, NY: Basic Books. Boyd, m., & Heer, J. (2006). Profiles as conversa- Goffman, E. (1959/1997). Self-presentation. In tion: Networked identity performance on Friend- C. Lemert & A. Branaman (Eds.), The Goffman ster. Paper presented at the Hawai’i International Reader. Malden, MA: Blackwell. Conference on System Sciences (HICSS-39), Kauai, HI. Golbeck, J. (2007). The dynamics of Web-based social networks: Membership, relationships, and Chattopadhyay, P., George, E., & Lawrence, S. change. First Monday, 12(11), 1–15. (2004). Why does dissimilarity matter? Explor- ing self-categorization, self-enhancement, and Gross, B. M. (2004). Multiple email addresses: uncertainty reduction. The Journal of Applied A socio-technical investigation. Paper presented Psychology, 89(5), 892–900. doi:10.1037/0021- at the First Conference on E-mail and Anti-Spam 9010.89.5.892 (CEAS), Mountain View, CA. Derlega, V. J., & Chaikin, A. L. (1977). Privacy Haney,C.,Banks,W.C.,&Zimbardo,P.G.(1973). and self-disclosure in social relationships. The A study of prisoners and guards in a simulated Journal of Social Issues, 33(3), 102–115. prison. Naval Research Review, 30, 4–17. Donath, J. (1999). Identity and deception in the Heer, J., & boyd, D. (2008). Visualizing online virtual community. In M. A. Smith & P. Kollock social networks. Retrieved July 11, 2008 from (Eds.), Communities in cyberspace (pp. 29-59). http://jheer.org/vizster/ London, UK: Routledge. Hogg, M. A. (2001). Social categorization, de- Erikson, E. H. (1963). Childhoodandsociety(2nd personalization, and group behavior. In M. A. Ed.). New York, NY: Norton. Hogg & S. Tinsdale (Eds.), Blackwell handbook of social psychology: Group processes (pp. 57- Erikson, E. H. (1980). Identity, youth, and crisis. 85). Malden, MA: Blackwell. In. New York, NY: Norton.

1358 Social Identities, Group Formation, and the Analysis of Online Communities

Hogg, M. A., Cooper-Shaw, L., & Holzworth, Kossinets, G., & Watts, D. (2006). Empirical D. W. (1993). Group prototypicality and deper- analysis of an evolving social network. Science, sonalized attraction in small interactive groups. 311, 88–90. doi:10.1126/science.1116869 Personality and Social Psychology Bulletin, 19(4), Lea, M., & Spears, R. (1995). Love at first byte? 452–465. doi:10.1177/0146167293194010 Building personal relationships over computer Hogg, M. A., & Hains, S. C. (1996). Intergroup networks. In J. T. Wood & S. Duck (Eds.), Un- relations and group solidarity: Effects of group derstudied relationships: Off thebeaten track(pp. identification and social beliefs on depersonal- 197-233). Thousand Oaks, CA: Sage. ized attraction. Journal of Personality and Social Linden Research Inc. (2008). About Second Life. Psychology, 70(2), 295–309. doi:10.1037/0022- Retrieved April 21, 2008, from http://secondlife. 3514.70.2.295 com/ Hogg, M. A., & McGarty, C. (1990). Self-catego- Lynch, D. (2005). Children’s identity develop- rization and social identity. In D. Abrams & M. A. ment in virtual spaces. Unpublished Dissertation, Hogg (Eds.), Social identity theory: Constructive McGill University, Montreal, QB. and critical advances (pp. 10-27). New York, NY: Harvester Wheatsheath. Marques, J. M., Abrams, D., Paez, D., & Hogg, M. A. (2001). Social categorization, social identi- Hogg, M. A., Terry, D., & White, K. M. (1995). fication, and rejection of deviant group members. A tale of two theories: A critical comparison In M. A. Hogg & S. Tinsdale (Eds.), Blackwell of identity theory with social identity theory. handbook of social psychology: Group processes Social Psychology Quarterly, 58(4), 255–269. (pp. 400-424). Malden, MA: Blackwell. doi:10.2307/2787127 McKenna, K. Y. A., Green, A. A., & Gleason, James, W. (1891/1950). The consciousness of M. J. (2002). Relationship formation on the self. In Principles of Psychology. New York, NY: Internet: What’s the big attraction? The Journal Dover Publications. of Social Issues, 58(1), 9–31. doi:10.1111/1540- Jenkins,R.(2000).Categorization:Identity,social 4560.00246 process and epistemology. Current Sociology, 48(3), McPherson, M., Smith-Lovin, L., & Cook, J. 7–25. doi:10.1177/0011392100048003003 (2001). Birds of a feather: Homophily in social net- Jenkins, R. (2004). Social Identity. New York, works. Annual Review of Sociology, 27, 415–444. NY: Routledge. doi:10.1146/annurev.soc.27.1.415 Jones,S.,&Madden,M.(2002).TheInternetgoes Munroe, R. L. (2000). Ethnography. In A. E. to college: How students are living in the future Kadzin (Ed.), Encyclopedia of Psychology (Vol. with today’s technology [Electronic Version]. 3, pp. 267-269). Washington, DC: American Pew Internet & American Life Project. Retrieved Psychological Association. October 15, 2007, from http://www.pewinternet. MySpace,Inc.(2008).AboutMySpace.Retrieved org/pdfs/PIP_College_Report.pdf April 21, 2008, from http://www.myspace.com/ Korte, R. (2007). Areview of socialidentitytheory index.cfm?fuseaction=misc.aboutus with implications for training and development. Journal of European Industrial Training, 31(3), 166–180. doi:10.1108/03090590710739250

1359 Social Identities, Group Formation, and the Analysis of Online Communities

Postmes, T., Spears, R., & Lea, M. (1998). Turner, J. C. (1987). Aself-categorization theory. Breaching or building social boundaries? SIDE- In J. C. Turner, M. A. Hogg, S. D. Oakes, S. D. effects of computer-mediated communication. Reicher & M. S. Wetherell (Eds.), Rediscovering Communication Research, 25(6), 689–715. the social group: Aself-categorization theory(pp. doi:10.1177/009365098025006006 42-67). Oxford, UK: Basil Blackwell. Postmes, T., Spears, R., & Lea, M. (2002). In- Tyler, T. R. (2002). Is the Internet changing social tergroup differentiation in computer-mediated life? It seems the more things change, the more communication: Effects of depersonalization. they stay the same. The Journal of Social Issues, Group Dynamics, 6(1), 3–16. doi:10.1037/1089- 58(1), 195–205. doi:10.1111/1540-4560.00256 2699.6.1.3 Valkenburg, P. M., Schouten, A. P., & Peter, J. Rubin,A.(1975).Disclosingoneselftoastranger: (2005). Adolescents’identity experiments on the Reciprocity and itslimits. JournalofExperimental Internet. New Media & Society, 7(3), 383–402. Social Psychology, 11, 233–260. doi:10.1016/ doi:10.1177/1461444805052282 S0022-1031(75)80025-4 Wallace, P. (1999). ThepsychologyoftheInternet. Ruitenberg, C. W. (2003). From designer identi- Cambridge, UK: Cambridge University Press. ties to identity by design: Education for identity Walther, J. B. (2007). Selective self-presentation de/construction [Electronic Version]. Philosophy in computer-mediated communication: Hyperper- of Education 2003. Retrieved October 12, 2007, sonal dimensions of technology, language, and from http://www.ed.uiuc.edu/EPS/PES-Year- cognition. Computers in Human Behavior, 23(5), book/2003/ruitenberg.pdf 2538–2557. doi:10.1016/j.chb.2006.05.002 Tajfel, H. (1974). Social identity and intergroup Wellman, B. (2000). Social network analysis: behavior. Social Sciences Information. Infor- Concepts, applications, and methods. In A. E. mation Sur les Sciences Sociales, 13, 65–93. Kadzin (Ed.), Encyclopedia of Psychology,7, 351- doi:10.1177/053901847401300204 352. Washington, DC: American Psychological Tajfel, H. (1981). Human groups and social cat- Association. egories: Studiesin social psychology.Cambridge, Wexelblat, A., & Maes, P. (1999). Footprints: UK: Cambridge University Press. History-rich tools for information foraging.Paper Taylor, J., & MacDonald, J. (2002). The effects of presented at the Conference on Human Factors in asynchronous computer-mediated group interac- Computing Systems, Pittsburgh, PA. tion on group processes. SocialScienceComputer Zimbardo, P. G. (1969). The human choice: Indi- Review, 20(3), 260–274. viduation, reason and order vs. deindividuation, Turner, J. C. (1985). Social categorization and the impulse and chaos. In W. J. Arnold & D. Levine self-concept: A social cognitive theory of group (Eds.), Nebraska Symposium on Motivation (Vol. behavior. In E. J. Lawler (Ed.), AdvancesinGroup 17, pp. 237-307). Lincoln, NE: University of Processes (Vol. 2). Greenwich, CN: JAI Press. Nebraska Press.

1360 Social Identities, Group Formation, and the Analysis of Online Communities

KEY TERMS AND DEFINITIONS experienceofthenominalidentity.Inotherwords, your nominal identity is what you believe you are Artifacts of Digital Performance: Artifacts (internal dialectic), and your virtual identity is the of digital performance refer to traces of interac- experience of being (external dialectic). tion history (Wexelblat & Maes, 1999), such as Self-Categorization: Self-categorization previous discussion postings and posted images, theory (Turner, 1985, 1987) suggests that iden- that new network members use as virtual cues to tification with any group is based on the extent interpret and build social context. to which individuals can enhance their social Collective Identity: Collective identification identitythroughcategorizingthemselvesasgroup is a representation of how people are similar to members (Chattopadhyay et al., 2004). each other based on the psychological connec- Social Identity: Social identity is central in tion between the self and social group (Abrams understanding intergroup relations and is the key & Hogg, 2001; Jenkins, 2004). element linking an individual to his or her social Depersonalization:Depersonalizationcauses group (Tajfel, 1974, 1981). people to conform to the group prototype and Social Network Analysis: Social network behave according to group norms. analysis involves the theorizing, model building Ethnography: Ethnography is a method of and empirical research focused on uncovering the research primarily concerned with the description patterning of links among network members (Free- of natural human communities (Munroe, 2000) man, 2000). Social network analysis conceives of and enables the interpretation of the flow of social social structure as a social network: a set of social discourse (Gertz, 1973/2000). actors and a set of relations ties connecting pairs Nominal and Virtual Identity: A nominal of these actors (Wellman, 2000). identity is the label with which an individual is identified and a virtual identity is an individual’s

This work was previously published in Handbook of Research on Social Software and Developing Community Ontologies, edited by S. Hatzipanagos; S. Warburton, pp. 86-101, copyright 2009 by Information Science Reference (an imprint of IGI Global).

1361