Socialization of Social Anxiety in Adolescent Crowds

J Abnorm Child Psychol (2011) 39:1239–1249 DOI 10.1007/s10802-011-9533-3 Socialization of Social Anxiety in Adolescent Crowds Nejra Van Zalk & Maarten Herman Walter Van Zalk & Margaret Kerr Published online: 22 June 2011 # Springer Science+Business Media, LLC 2011 Abstract In this study, we looked at whether social anxiety is socialized, or influenced by peers’ social anxiety, more in some peer crowds than others. Adolescents in crowds with eye-catching appearances such as Goths and Punks (here termed Radical), were compared with three comparison groups. Using data from 796 adolescents (353 girls and 443 boys; Mage =13.36) at three timepoints, the results show that adolescents affiliating with the Radical crowd tended to select peers from the same crowd group. Being a member of a crowd in itself did not predict socialization of social anxiety, but adolescents in the Radical crowd were more influenced by their peers’ social anxiety than adolescents who did not affiliate with the Radical crowd group. The results suggest that through a bidirectional process, adolescents affiliating with Radical crowds may narrow their peer relationship ties in time, and in turn socialize each other’s social anxiety. Keywords Social anxiety . Peer crowds . Selection . Socialization . Early adolescence During adolescence, most young people seek an identity of their own. Part of this process involves affiliating with peers whose values and interests are compatible with their own (Newman and Newman 1976). Many young people adopt identities that separate them from others, but make them similar to one another. Such constellations of youths are sometimes categorized as crowds. Crowds are large groups of individuals, who are similarly stereotyped by N. Van Zalk (*) : M. H. W. Van Zalk : M. Kerr Center for Developmental Research, JPS:Psychology, Örebro University, Örebro, Sweden e-mail: nejra.van-zalk@oru.se themselves and others, and may or may not be friends (Brown 1990). Even though crowd members do not have to be friends in order to be assigned by themselves or others to a certain crowd, studies indicate that adolescents tend to have relationships with peers who endorse a similar, or the same crowd identity (Brown 1990; La Greca et al. 2001; Urberg et al. 2000). Crowds are important for many young people, as they provide a sense of identity and prospects for social interactions with other similar-minded adolescents (Brown 1990). Hence, affiliating with peer crowds might serve as a gateway to gaining an identity in adolescence, and might fulfill different needs for adolescents. Previous research has shown that crowd members tend to take on similar crowd characteristics such as values, attitudes, and clothing styles. The majority of research on similarity of crowd characteristics has identified typical crowds such as the higher-status Jocks and Cheerleaders, and the lower-status Nerds, Brains, Druggies, and Burnouts (see eg., Heaven et al. 2005; La Greca et al. 2001; Prinstein and La Greca 2002; Stone and Brown 1998; Thurlow 2001; Urberg et al. 2000). Affiliating with high-status crowds in adolescence has been linked to higher self-esteem and lower depression (Brown and Lohr 1987; Brown et al. 2008; Prinstein and La Greca 2002), whereas affiliating with low social-status crowds has been shown to result in increases in social anxiety, depression, and loneliness (Prinstein and La Greca 2002). Crowd affiliation has also been linked to health-risk behavior (Kobus 2003; La Greca et al. 2001), as well as eating, exercise, and weight control behaviors (Mackey and La Greca 2007). How can this similarity be explained? There are a number of possible reasons for similarity between adolescents and their peers. According to the similarity-attraction hypothesis, people are attracted to others who are similar to themselves because this enhances 1240 interaction between them (Byrne 1971), and this selection is one reason for similarity. Similarity, however, could also be a result of influence, or socialization (Kandel 1978), a process in which individuals become similar over time because they influence each other. Concerning crowds, there is evidence for selection into peer crowds, but less is known about socialization within peer crowds. Previous research indicates that crowd members are likely to select each other as friends, as 55–85% of adolescents befriend someone from their own or a similar crowd (Brown 1990; La Greca et al. 2001; Urberg et al. 2000), and adolescents are most friendly to members of their own crowd (Urberg et al. 2000). Although socialization has not been studied, we suggest that this preference for similar peers within crowd contexts could be a basis for socialization, as adolescents might have opportunities to model behavior or even take direct instructions from crowd members on how to behave. In addition, peer crowds are thought to comprise cliques, or tight-knit groups of a smaller number of adolescents (Brown 1990). As adolescents interact with each other within these contexts as well, cliques—especially if embedded in crowds—may serve as an additional context for socialization of behaviors. Through possibly intense social interactions with other clique members, adolescents might encourage or model behaviors from each other. Hence, there are some reasons to expect socialization within peer crowds, and this could be an important issue, because some of the characteristics on which crowd members are similar concern risk behaviors and internal adjustment. Similarities on internal adjustment have been shown for alternative crowds such as Punks, Goths, and Metal Heads (Bešić and Kerr 2009; Delsing et al. 2007). For example, one study focused on shy, socially fearful behavior in peer crowds with extremely eye-catching, or even shocking appearances such as Punks and Goths (termed Radical). In these crowds adolescents might, for example, wear white face paint, faked bloodstains around the eyes or mouth, or long, black hair shaved off of one side of the head. The authors hypothesized that creating the unusual appearances that define these crowds might be a way of limiting anxiety-arousing social contacts or a self-handicapping strategy, where social failures could be ascribed to one’s extreme appearance rather than one’s social anxiety (Bešić and Kerr 2009). Radicals were compared with three other groups. The first one, Academics, consisted of Computer Nerds and Brains, which are often considered by their peers to be more socially anxious than adolescents in other crowds (Kinney 1993). The second, the Independents, consisted of adolescents who answered “None of the above [identities]” to the crowd question. These adolescents have previously been found to have high levels of social anxiety (Prinstein and La Greca 2002). These two groups were used J Abnorm Child Psychol (2011) 39:1239–1249 in order to compare the Radicals with crowds that were known to be relatively high on social anxiety, but without the startling appearances of the Radicals. The final comparison group comprised crowds that might be recognizable by their appearances, but whose appearances were not deemed shocking or off-putting, for example “Sports,” “Role-Players,” “Vegans,” and “Skaters.” The socially anxious adolescents in the Radical crowd group were found to be more socially anxious and depressed and to have worse self-esteem than socially anxious adolescents in all other crowd groups. The study was cross-sectional, however, so there was no way of teasing apart selection and socialization effects. Thus, whether social anxiety socialization occurs after adolescents join these crowds remains untested. Very little is known about whether social anxiety might be socialized in Radical crowds more than other crowds. Socially anxious individuals tend to conform more easily than others (Cohen and Prinstein 2006; Santee and Maslach 1982). Scholars have suggested that conformity might be a vital part of being in a crowd that maintains a specific clothing style, and that being in one of these crowds might increase adolescents’ susceptibility to peer influence (Brown et al. 1990). Going beyond clothing, the Radical crowds, more than others, might value a reserved, withdrawn, or even seemingly depressed way of behaving, and adolescents in these crowds might intensify their socially fearful behavior and/or depressed mood to conform to the standard. Then, through self-perception (Bem 1967), a process in which people infer their own attitudes or characteristics from their behaviors, they might come to see themselves as increasingly socially anxious. In addition, even though crowds in themselves might be abstract phenomena, they do affect adolescents’ social relationships (Brown et al. 1994), as adolescents are more likely to be friends if they are part of the same crowd (Brown et al. 1994; La Greca et al. 2001; Urberg et al. 2000), and peers influence each other on both delinquency (Burk et al. 2007; Dishion et al. 1996; Kiesner et al. 2002) and depression (Hogue and Steinberg 1995; Prinstein 2007; Rose 2002; Van Zalk et al. 2010). As Radical crowds have been shown to comprise fewer individuals than other crowds (Bešić and Kerr 2009), one could assume that these adolescents might have more tight-knit, close peer relationships than adolescents in larger crowds. If so, adolescents in the Radical crowds may run a greater risk than adolescents in other, larger crowds of being influenced by their peers on social anxiety. Previous research supports this notion, showing that socially anxious girls in close peer relationships, albeit not peer crowds, were at higher risk of peer contagion of social anxiety than less socially anxious girls (Prinstein 2007). To our knowledge, however, no studies have looked at differential socialization of social anxiety within peer crowds in adolescence. J Abnorm Child Psychol (2011) 39:1239–1249 1241 An important factor to take into account regarding socialization of social anxiety within contexts of peer crowds are gender differences. First, differences on levels of social anxiety between boys and girls are well documented (e.g., La Greca and Lopez 1998), and scholars have also suggested that social anxiety may be more socially accepted in girls than in boys (Bruch and Cheek 1995). This may make girls’ social fears more prone to influence than boys’. Some studies support this notion, as socially anxious girls have been found to be more susceptible than boys to influence from their best friends on depression (Prinstein 2007). Second, girls are more likely to be part of friendship cliques (Urberg et al. 2000), whereas boys tend to have larger peer groups (Benenson 1990). In addition, adolescent girls’ relationships with peers are characterized by high levels of psychological intimacy (Selfhout et al. 2009). Even though this process is assumed to function on the dyadic level, it may have relevant implications for why peer crowds would influence the development of social anxiety in girls more than boys. Thus, as we may expect different outcomes for boys and girls regarding socialization of social anxiety, it is of consequence to consider possible gender differences in these processes. In this study, we examined the socialization of social anxiety within peer crowds in early adolescence, using longitudinal data on early adolescents. We controlled for depression and self-esteem in the analyses, as depression (Elovainio et al. 2004; Smith and Betz 2002; Stednitz and Epkins 2006) and poor self-esteem (Crozier 1981; Pilkonis 1977) are well-known correlates of social anxiety, and should be controlled to determine the unique role of social anxiety in adolescent peer relationships (Romero and Epkins 2008). In so doing, we excluded the likelihood of depression and self-esteem being reasons for selection and socialization processes on social anxiety. We hypothesized, first, that adolescents would select peers based on social anxiety. Second, we hypothesized that adolescents would also select peers based on crowd membership, and that adolescents affiliating with Radical crowds would have relationships with peers who affiliate with Radical crowds. Third, we hypothesized that peers would socialize each other’s social anxiety more in the Radical crowds than in other crowd groups. Finally, we hypothesized that social anxiety may be socialized among peers in the crowds differently for boys and girls, in that girls in Radical crowds may be more strongly influenced by peers’ social anxiety than boys. 26,000 in Sweden. The data collections took place annually at roughly one-year intervals, with five waves of data collected. More than 90% of all adolescents in the community who were in grades 4–12 participated (aged approximately 10 to 18 years). Measures of social anxiety, depression, and crowd affiliation were included at Waves 3, 4, and 5, so these waves were used in this study and referred to as Times 1, 2, and 3. At Wave 3 of the study, the unemployment rate (8%) was comparable to that in the rest of the country, as was the percentage of singleparent households for the entire community (10.0%). The mean individual income was around 21,000 USD, which was about 5.0% lower than in the rest of the country (Statistics Sweden 2004). To form the target sample that would be included in the analyses, we began with all 7th-graders who participated in at least two consecutive measurements (N=329; 148 girls, and 181 boys; Mage =13.27). After inclusion of the peers these 7th-graders nominated at each time point, the final network consisted of 796 adolescents (353 girls and 443 boys; M = 13.36). The peer relationship nominations included both in- and out-of-school peers, and several adolescents named older or younger peers. Hence, the final social network included 4th through 12th graders, with adolescents spread over 27 classes and 4 different schools. Two and a half percent of the target sample were 4th-graders (about 10 years old), 6% were 5th-graders, 14.7% were 6th-graders, 44.3% were 7th-graders, 17.3% were 8th-graders, 6.9% were 9th-graders, 2.3% were 10thgraders, 0.4% were 11th-graders, and 0.1% were 12th-graders. Regarding the overall missingness of the peer relationship variables in the target sample, all of the target participants had at least one peer nomination at one timepoint. Regarding missingness for other variables (e.g., social anxiety), the program we used imputes or estimates information about covariates based on other existing participant data (Huisman and Snijders 2003). Eighty-nine percent of the adolescents in the sample completed 1 measurement, 78.3% completed 2 measurements, whereas 65.7% completed all 3 measurements of data on social anxiety. Roughly 4% of the adolescents in the target sample were first-generation immigrants at Time 1. Sixty-eight percent of the sample lived in households with both biological parents; 19% lived in single-parent households, and 13% lived with one stepparent and one biological parent. Method Procedure Participants Adolescents filled out questionnaires during school time. They were informed about the types of questions they would answer, and how long it would take to answer them. They were told that participation in the study was voluntary, Participants were from a large, community-based study conducted in a community with a total population of about 1242 and that they might choose to do something else should they decide not to participate. They were also guaranteed that their answers would be strictly confidential, and never shown to parents, teachers, or other people. Trained research assistants administered the questionnaires, with no teachers present during the sessions. Before the data collections, parents were informed about the study through community-based meetings and via letters. With the letter, they received a postage-paid card to return without cost if they did not want their adolescent to participate (only 1% of the parents did so). The parents were also informed that they could withdraw their adolescent from the study at any time. No participant was paid for taking part in the study. In each of the classes we held a drawing for movie tickets, however, for which all adolescents were eligible, regardless of participation. All of the measures and procedures were approved by the University Ethics Review Board. Measures Peer Relationships Adolescents were asked to identify up to “three important peers, someone you talk with, hang out with, and do things with” (Kerr et al. 2007). In addition, they identified up to 10 peers with whom they spent time in school and up to 10 peers with whom they spent time outside of school. Up to 23 reciprocal and nonreciprocal nominations of peers each participant spent time with in school and in free-time were collected annually. That is, adolescents were allowed to nominate peers regardless of whether these peers reciprocated the peer relationship or not. For the current study, when participants nominated siblings or romantic partners as important peers, these were excluded from the analyses so that we could focus only on other peers. We combined the three different types of peer relationship nominations by identifying the unique peers who were nominated at each measurement. The mean number of nominated peers ranged between 3.2 and 4.6 across the measurements. Of all peer relationships in the network, 36–41% were formed from one measurement to the next, 26–31% were ended from one measurement to the next, and 30–33% remained stable from one measurement to the next. Self-Perceptions of Crowd Affiliation Before developing a peer crowd affiliation measure, we collected information about existing crowds by talking to teachers, students, and adolescent leaders in the community. We then developed a list of peer crowds that, based on the information we received, existed in the community. This list was included in the questionnaires, and adolescents were asked to choose the peer crowd they most identified J Abnorm Child Psychol (2011) 39:1239–1249 with from the list. They were also given the option to indicate that they did not identify with any crowd, or they could add a crowd if the one they identified with was not on the list. Formation of Crowd Groups We combined crowds into crowd groups in the same way as was done in a previous study that examined Radical crowds and socially anxious behavior (Bešić and Kerr 2009). The idea was to compare adolescents in the Radical crowds with those in other crowds that have been shown to be high on social anxiety, thus creating a stringent test of the role of Radical crowds in the socialization of social anxiety. The Radical Crowd Group Previously, several crowds were combined into a Radical crowd, and the predictive validity of the measure was shown (Bešić and Kerr 2009). Peer crowds that were considered by those who had spent time in the schools teaching or collecting data as eye-catching in a way that seemed calculated to shock, scare, or put people off were included in the Radical crowd group. The Punks and Goths appear in other studies of peer crowds, whereas the Synths (so named for their interest in electronically synthesized music, and their all-black clothing and oftenblack, dramatically styled hair) and the Aesthetics may be specific to adolescent culture in Sweden. The Aesthetics create startling appearances in a large variety of ways, for example, creating space-alien-type hair sculptures or wearing welder’s goggles, and are typically found among adolescents who have chosen or are planning to choose the music, drama, and art high school track. Thus, the Punks, Goths, Synths, and Aesthetics formed the Radical crowd group. Fourteen percent affiliated with the Radicals across all three timepoints; 29% remained affiliated with the Radicals from Time 1 to Time 2, and 23% remained affiliated with the Radicals from Time 2 to Time 3. Thus, Radical crowd membership was somewhat stable over time. Comparison Crowd Groups As in a previous study (Bešić and Kerr 2009), and as explained in the introduction, we compared Radicals with three crowd groups: Academics, Independents, and Non-Radicals. Table 1 shows the numbers of participants in each crowd for Times 1–3, as well as the entire list of crowds. Social Anxiety Social anxiety was measured with 8 questions about fears in different situations (Gren-Landell et al. 2009). Examples of these were speaking in front of the class, putting a hand up during class, and making a phone call to someone one does not know. The adolescents rated themselves on a three-point scale as having No fear (1) to A lot of fear (3) in these situations. The Cronbach’s alphas J Abnorm Child Psychol (2011) 39:1239–1249 1243 Table 1 Numbers and ranges of participants in all crowd groups by gender at times 1–3 Crowd groups Self-perceptions of crowd affiliation Radicals (NT1-T3 =14–45) Aesthetics Goths Punks Synths None of the Identities Computer Nerds Brains Sports Hip-Hoppers Role-Players Snobs Poppers Skaters Feminists Vegans Ravers Environmental Activists Independents (NT1-T3 =111–136) Academics (NT1-T3 =66–128) Non-Radicals (NT1-T3 =257–363) Girls (N=238) Time 1 Boys (N=260) Girls (N=268) Time 2 Boys (N=357) Girls (N=225) Time 3 Boys (N=281) 2 2 5 0 98 8 6 85 9 5 5 4 3 4 1 0 1 2 2 1 0 38 46 6 117 17 10 4 4 7 0 1 4 1 11 2 4 0 59 24 24 76 15 6 20 19 0 5 1 0 2 3 6 10 3 40 61 19 132 19 14 16 9 12 3 3 4 3 12 2 4 1 71 13 12 70 12 7 11 7 2 0 0 1 0 5 8 10 3 40 61 7 82 11 12 16 3 8 4 3 5 3 Total N: were 0.73 for Time 1, 0.71 for Time 2, and 0.76 for Time 3. The cross-year correlations ranged from 0.34 to 0.60. Depression Depression was measured with the widely-used Child Depression Scale from the Center for Epidemiological Studies, which has shown good validity (Radloff 1977). There were 20 questions rated on four-point scales from Not at all (1) to Often (4). The scale assesses depressive symptoms such as worry, sadness, hopelessness, lethargy, and poor appetite. The adolescents were asked to think about the past week. Examples of items were: I have “worried about things I don’t usually worry about,” “felt scared,” and “felt down and unhappy.” The Cronbach’s alphas were 0.84 for Time 1, 0.83 for Time 2, and 0.84 for Time 3. The cross-year correlations ranged from 0.39 to 0.51. Correlations between depression and social anxiety ranged from 0.11 to 0.33 (ps<0.001–0.05) at the three time points. Self-Esteem Self-esteem was measured with the Rosenberg scale (Rosenberg 1979). There were ten items, assessing self-perceptions of one’s own worth. Examples of items were: “On the whole, you are satisfied with yourself,” “You think that you have many good characteristics,” and “Sometimes you feel you are of no use.” The four-point response scale ranged from Not true at all (1) to Very true (4). The Cronbach’s alphas were 0.82 for Time 1, 0.86 for Time 2, and 0.77 for Time 3. The cross-year correlations 498 625 506 ranged from 0.53 to 0.60. Correlations between self-esteem and social anxiety ranged from 0.15 to 0.32 (ps<0.001) at the three time points. Strategy for Analysis To examine selection and influence processes simultaneously, we used the software program Simulation Investigation for Empirical Network Analyses, or SIENA (Snijders 2001; Snijders et al. 2007). This method has been successfully applied to assess selection effects for personality in adolescence (Selfhout et al. 2010), and selection and influence processes regarding social anxiety (Van Zalk et al. 2011). SIENA is currently the only program available that can be used to study influence processes while controlling for selection within adolescents’ networks of multiple peer relationships. The analyses take into account that some peer relationships evolve into reciprocal peer relationships whereas others do not, and how all these peer relationships in turn affect adolescents over time. When socialization is studied, then, reciprocity is controlled for. Age, gender, internalizing, and whether adolescents selected others based on being in the same classroom and school were also controlled for in all of the subsequent analyses. If information about any of the variables in the current study were missing (>8% across measurements), scores were imputed according to the procedure developed by Huisman and Steglich (Huisman and Steglich 2008). More details about the modeling and 1244 J Abnorm Child Psychol (2011) 39:1239–1249 specific effects regarding this approach are provided in the results section. Results Control Variables Several effects of control variables on peer relationship selection were found, and are shown in Table 2. First, the negative outdegree parameter showed that participants were likely to nominate significantly less than 50% of the total network members as peers on average, thereby indicating that adolescents were selective when forming new peer relationships. Second, the reciprocated relationship effects, referring to the tendency to form reciprocated peer relationships, showed that adolescents tended to reciprocate their peer relationships. Third, the triadic relationship effects, referring to the tendency to form triadic peer relationships, indicated that adolescents tended to directly select the peers of their peers over time. Table 2 Selection effects within the social network Parameter Social anxiety Social anxiety: selection similarity Social anxiety: adolescents select Social anxiety: peers select Crowd group affiliation Radicals: selection similarity Radicals: adolescents select Radicals: peers select Academics: selection similarity Academics: adolescents select Academics: peers select Independents: selection similarity Independents: adolescents select Independents: peers select Non-Radicals: selection similarity Non-Radicals: adolescents select Non-Radicals: peers select Control variables Outdegree Reciprocated relationship Triadic relationship Gender: selection similarity Age: selection similarity Depression: selection similarity Self-esteem: selection similarity * p<0.05; ** p<0.01; *** p<0.001 Unstandardized effect s.e. 0.85*** 0.09*** −0.01* 0.30 0.01 <0.01 1.25* −1.39* −1.30* −0.09 0.13 −0.19*** 0.10 0.04 −0.10 0.13*** 0.23*** −0.06 0.60 0.62 0.60 0.07 0.07 0.07 0.06 0.07 0.06 0.04 0.04 0.04 −3.42*** 2.84*** 0.43*** 0.69*** 2.23*** 1.64*** 0.02 0.03 0.04 0.01 0.02 0.11 0.13 0.24 Thus, effects of the network structure on selection were significant and therefore needed to be controlled for in further analyses. In addition, all main effects were controlled for gender, age, depression, and self-esteem when looking at these similarity effects, as demonstrated in Table 2. Boys selected boys more than girls, and girls selected girls more than boys (as indicated by the gender: selection similarity effect). Furthermore, adolescents tended to select peers of the same age (as suggested by the age: selection similarity effect). Adolescents with similar levels of depressive symptoms tended to select each other as peers (depression: selection similarity). Finally, no significant effects of self-esteem on peer relationship selection were found, suggesting that mean levels of and similarity in self-esteem did not affect peer relationship selection. Selection of Peers Based on Social Anxiety For selection processes, we examined to what extent variables predicted that a non-peer dyad at one measurement would change into a peer dyad at the next. The parameters were calculated across all three measurements, and controlled for reciprocation and non-reciprocation of the dyad. This means that when examining changes in the dependent variable selection (selecting a person as a peer versus not selecting this person as a peer), the program calculated whether this selection was reciprocated or unreciprocated. The information from each wave was used for the prediction of the subsequent wave. To test whether similarity in social anxiety predicted peer selection, we examined to what extent similarity in social anxiety between all non-peers predicted selecting a peer or not selecting a peer. This effect, labeled social anxiety: selection similarity, is shown in Table 2. Higher social anxiety: selection similarity predicted a higher probability of peer selection over non-selection. Hence, adolescents tended to select peers with similar levels of social anxiety. Selection of Peers Based on Crowd Group Membership The Radical Crowd Group As the findings in Table 2 show, adolescents tended to select peers with the same Radical crowd membership (indicated by the Radicals: selection similarity effect). Adolescents from the Radical crowd group were more likely to select other Radical crowd group members, whereas the other adolescents were more likely to select peers who were not from the Radical crowd group. Adolescents from the Radical crowd group also tended to select fewer peers (indicated by Radicals: adolescents select), and tended to be selected less by others (indicated by Radicals: peers select). J Abnorm Child Psychol (2011) 39:1239–1249 The Academic Crowd Group Being in the Academic crowd group did not significantly predict peer selection (as shown by a non-significant Academics: selection similarity effect in Table 2). Adolescents in the Academic crowd group were less selected by others as peers (indicated by Academics: peers select), and the Academic crowd group membership (i.e., Academics: adolescents select) did not significantly predict how often they selected peers. Hence, Academics did not tend to select each other as peers, nor were they selected often by other adolescents in the network. The Independent Group Being in the Independent group did not significantly predict peer selection (as shown by Independents: selection similarity in Table 2). In a similar vein, adolescents’ Independent status did not predict how often they selected peers (i.e., Independents: adolescents select), or how often they were selected by others as peers (Independents: peers select). Thus, Independent status was not predictive of peer relationship selection. The Non-Radical Crowd Group For adolescents in all other crowds, the results showed that Non-Radicals: selection similarity predicted peer selection, implying that adolescents in Non-Radical crowds preferred peer relationships with other adolescents from Non-Radical crowds. Additionally, adolescents from the Non-Radical crowd group tended to select a larger number of peers (as shown by the significant effect for Non-Radicals: adolescents select). In addition, being from the Non-Radical crowd group did not affect being selected as a peer by others (indicated by NonRadicals: peers select). In sum then, even though NonRadicals tended to select each other as peers and tended to select a larger number of peers, they did not tend to be selected particularly often by others in the network. Socialization of Social Anxiety Examining Socialization of Social Anxiety For socialization processes, the dependent variable was represented by changes in adolescents’ social anxiety. SIENA treats changes in social anxiety as ordinal categorical values (Snijders et al. 2007) and models changes in peer relationships and social anxiety as two interdependent processes, thereby controlling each for the other. This means that positive significant effects of the variables of interest on these change scores of social anxiety indicated that higher scores on the variable in question predicted a larger increase in adolescents’ social anxiety than expected by chance, holding all other effects of independent variables (including those in selection) constant. When examining these effects, individuals are assumed to be mutually dependent because each individual figures simultaneously as an adolescent 1245 who is being influenced, and as a peer who influences others. This mutual dependence is mediated by effects of the network structure, such as reciprocated relationships and triadic relationships. These effects, respectively, pertain to the tendency to form reciprocated relationships (John selects Mark and Mark selects John) and the tendency to form triadic relationships (John selects Mark, Mark selects Sue, and John selects Sue). This also controls for the effect of small cliques on socialization, in that our results then do not pertain to small clique formations of up to three individuals. This way, all selection and influence effects were adjusted for overlap in the network (for full details, see Snijders 2001). We also controlled for the trend in social anxiety by including both the linear tendency of change and quadratic tendency of change in the categorical scores of social anxiety (for more details, see Snijders et al. 2010). Hence, prior levels of social anxiety were controlled for. We will discuss effects of targets‘ characteristics, peers’ characteristics, and interactions between effects. Effect of Peers’ Social Anxiety Our findings demonstrated that peers’ social anxiety significantly predicted changes in adolescents’ social anxiety, as shown in Table 3. Adolescents showed a tendency to become similar to their peers’ average social anxiety over time, thus supporting the idea of socialization processes in peer relationships for social anxiety. Was Social Anxiety Socialized among Peers in the Radical Crowd Group more than Others? To answer this, our primary question of the study, we looked at the interactions in the model. They are shown in Table 3. In particular, the Peers’ social anxiety X Targets’ Radical affiliation interaction indicates whether social anxiety was socialized more among peers in the Radical crowd group than in the other crowd groups. As shown in the table, this was the only significant interaction in the model, suggesting that the answer is yes. We explored the interaction by using post-hoc multiple group tests (Snijders et al. 2010) to compare effects for adolescents who affiliated with the Radical crowd group at at least one measurement (n=82) with those who never affiliated with the Radical crowd group (n=714). Figure 1 displays the differences between the Radical crowd groups (unstandardized effect=1.13, p<0.01) and the other crowd groups (unstandardized effect = 0.33, p < 0.01) in the socialization of socially anxious behavior over time. Although peers’ social anxiety predicted changes in adolescents’ social anxiety for both groups (unstandardized effect=0.87, p>0.05), the effect was significantly larger (unstandardized effect=1.13, p<0.05) for adolescents who affiliated with the Radical crowd group than for adolescents who did not. Thus, our 1246 Table 3 Socialization effects within the social network * p<0.05; ** p<0.01; *** p<0.001 J Abnorm Child Psychol (2011) 39:1239–1249 Parameter Linear shape social anxiety Quadratic shape social anxiety Peers’ social anxiety Peers’ social anxiety * Targets’ Peers’ social anxiety * Targets’ Peers’ social anxiety * Targets’ Peers’ social anxiety * Targets’ Radical affiliation Non-Radical affiliation Academic affiliation Independent affiliation findings suggest that adolescents in the Radical crowd group were more influenced by their peers’ social anxiety than adolescents who were not members of the Radical crowd group. Is Social Anxiety Socialized Differently for Boys and Girls among Peers in the Crowds? To answer this question, we computed a three-way interaction between gender, Radical crowd affiliation, and friend’s social anxiety predicting changes in social anxiety. We controlled for all other main effects, as well as two-way interaction effects (Gender X Radical Crowd; Gender X Peers’ Social Anxiety, and Radical Crowd X Social Anxiety) in the model. The interaction between gender, Radical crowd, and friends’ social anxiety when predicting target’s changes in social anxiety was not significant (unstandardized effect=0.01, p>0.10). This implies that there were no gender differences regarding the extent to which Radical crowd affiliation interacted with friend’s social anxiety when predicting targets’ social anxiety. Fig. 1 Interaction effects of Radical crowd affiliation by peers’ social anxiety on adolescents’ social anxiety. Change in social anxiety was controlled for previous levels of social anxiety Unstandardized effect s.e. −0.24*** −0.02 0.87*** 0.58*** 0.04 0.06 0.19 0.05 <0.01 0.23 0.06 0.04 0.05 0.58 Discussion Adolescents tend to affiliate with reputation-based collectives called peer crowds, and they are likely to form peer ties with others who affiliate with similar crowds. Previous research suggests, however, that membership in some peer crowds is linked to depression (Hogue and Steinberg 1995; Prinstein 2007; Rose 2002), and membership in crowds that adopt eye-catching, sometimes extreme clothing, hair, and makeup styles, in particular, is linked to social anxiety, depression, and low self-esteem (Bešić and Kerr 2009). In this study, we expanded on that knowledge by showing that, controlling for depression and self-esteem, adolescents with Radical crowd affiliation were more influenced by their peers’ social anxiety than were adolescents in other crowds. We also showed that peers within the Radical crowds became more and more isolated from the rest of the social network, as they not only selected others as peers to a lesser degree over time, but were also being selected less over time. This might point to a bidirectional process, where adolescents start affiliating with Radical crowds, narrow their peer relationship ties in time, and become less popular in the entire peer relationship network. Over time, they may in turn socialize each other into becoming more socially anxious, and these processes seemed to function similarly for boys and girls. To our knowledge, this is the first study to examine socialization processes of socially anxious behavior within crowds in early adolescence. There are several possible ways in which adolescents might socialize each others’ socially anxious behaviors. First, peers could strengthen one another’s social anxiety by signaling empathy or approval in situations that boost social fears, or via discussions of awkward, fearful feelings. A similar process, termed deviancy training, has been identified in delinquent adolescents’ peer relationships (Dishion et al. 1996). Talking or commenting about antisocial behavior has been linked to subsequent increases in the actual antisocial behaviors, and this process might be extended to social anxiety. As peer relationships are often embedded in crowds, this may be a viable explanation. J Abnorm Child Psychol (2011) 39:1239–1249 A second way in which adolescents’ socially anxious behaviors could be influenced by their peers is through dwelling on the problems of social anxiety in discussions with peers. This process, called co-rumination, has been linked to peer socialization of depression (Rose 2002). It refers to extensive discussions and focus on negative emotions in problems with close reciprocal friends. By strengthening the peer relationship bonds in dyads through extensive self-disclosure, co-rumination increases feelings of depression and anxiety (Rose 2002). Both depression and anxiety are well-known correlates of social anxiety (Cox et al. 2005; Elovainio et al. 2004), so co-rumination might also apply to social anxiety. Further, even though co-rumination is assumed to occur in close peer relationships, such interactions might also exist within the larger contexts of peer crowds (Brown 1990; La Greca et al. 2001; Urberg et al. 2000). Regarding socially anxious peers then, they might evoke even more socially fearful behaviors over time due to frequent, intense discussions about their social fears. A third way in which adolescents’ socially anxious behaviors could be affected by their peers‘ social anxiety is through modeling behavior to garner approval. Adolescents might model peers’ taciturn or withdrawn behaviors in order to gain status or acceptance in the peer group, and those who are already socially anxious might be more prone to doing so. In one study, socially anxious girls were found to be at higher risk for developing depressive symptoms over time due to a greater susceptibility to peer contagion (Prinstein 2007). As socially anxious individuals usually have poor self-esteem (McCarroll et al. 2009), they might be more vulnerable to peer contagion―both in close relationships and in the crowd social network. Consequently, socially anxious adolescents may be at increased risk to reinforce their social anxiety because they could be more vulnerable to peer contagion than less socially anxious adolescents. Extending this reasoning one step further, one may wonder why social anxiety would be socialized more in the Radical crowds than in others. One idea suggested in the literature is that conformity may be an important part of affiliating with a crowd that maintains a specific style of clothing, hair, and makeup, and this emphasis on conformity might increase adolescents’ susceptibility to peer influence (Brown et al. 1990). As previous research has shown, socially anxious adolescents tend to conform more easily than others (Cohen and Prinstein 2006; Santee and Maslach 1982). As was suggested previously, adolescents affiliating with the Radical crowds, more than others, could value taciturn, withdrawn social interactional styles and behaviors in addition to their eye-catching appearances. Adolescents in these crowds might intensify their socially anxious behaviors to conform to the group standard, 1247 without necessarily being aware of it. This might lead to perceptions of the self as more socially anxious over time, through a process of observing one’s own behavior and concluding what attitudes or characteristics must have caused it (Bem 1967). Another explanation for social anxiety being socialized more in the Radical crowds than in other crowds might lie in the way adolescents’ peer relationships are constructed. Studies show that adolescents are likely to have peers within the same crowd (Brown et al. 1994; La Greca et al. 2001). Radical crowds tend to be small (Bešić and Kerr 2009), and adolescents in these crowds might have more intimate, close peer relationships than adolescents in larger crowds. As close peers tend to influence each other on both externalizing (e.g., delinquency, see Burk et al. 2007; Dishion et al. 1996; Kiesner et al. 2002) and internalizing behavior (e.g., depression, see Hogue and Steinberg 1995; Prinstein 2007; Rose 2002), Radical adolescents might be more influenced by their peers’ social anxiety because their relationships may be closer than those in larger crowds. Because peer relationships in this study were partly measured in terms of important peers in adolescents’ lives, this seems a viable explanation. This study is not without limitations. First, we only included adolescents in our data if they had two consecutive waves of data with peer relationship nominations to fulfill the criteria of the SIENA software. Thus, adolescents who did not nominate any peers, which might indicate social isolation, were not part of the study. Hence, it remains unanswered whether they were affected by their peers. In addition, whether adolescents had stable or intimate, high-quality peer relationships was not examined in this study and is another limitation. Previous research shows, however, that the stability of peer relationships might not be the most important factor in children’s adjustment, but rather acquiring a peer (Laursen et al. 2007; Wojslawowicz Bowker et al. 2006). Peer relationships during adolescence also show high instability across short periods of time (Cairns et al. 1995). Also, we cannot rule out the possibility that our results might have been due to clique formations rather than crowds, as cliques are often embedded in peer crowds (Brown 1990). The size of typical cliques has been disputed, with a lack of consistency between how many individuals constitutes a clique (Henrich et al. 2000). In our study, nonetheless, we controlled for the effect of dyadic and triadic peer relationship structures, implying that the results were true over and above smaller clique formations. A further limitation is that we did not have peer reports of crowd affiliation, in order to verify adolescents’ self-perceptions of crowd membership, which is nonetheless a limitation shared by most previous studies. Finally, we concentrated on one cohort of 7th-grade adolescents (about age 13) and 1248 their peers. Perhaps adolescents are more open to socialization during this period than others, but we cannot say. Despite these limitations, there are several strengths of this study, as well. By using SIENA, we were able to control for various selection effects on adolescents’ peer relationships and various network effects, which have been shown to be important (Haynie 2001). Thus, our results provide unique insight into the selection and socialization effects of social anxiety, over and above other important effects. In addition, we included an entire cohort of adolescents in one community, followed over three years, and they were allowed to select peers both in and outside of their classrooms and schools. This provides an ecologically valid view of adolescents’ peer relationships. Finally, the reports of peers’ characteristics were independent of the target adolescents, as we included peers’ independent reports of their own behavior. This gives us confidence that our findings are not biased by adolescents’ projecting their own behavior onto their peers (see Bauman and Ennett 1994) or believing that their peers are more similar to themselves than they actually are (Prinstein and Wang 2005). Thus, despite its limitations, the study provides a new way of viewing socialization of socially anxious behaviors within adolescent crowds. Developing a sense of identity is one of the major developmental tasks for most young people. During some part of their teenage years, many adolescents explore their identities by adopting different ways of looking and behaving, and by affiliating with different crowds. For a smaller number of adolescents this may be taken to the extreme, as they might adopt startling or even shocking appearances. Besides being influenced by these crowds in terms of clothing or music, however, it seems that some behaviors may be socialized as well. Our data showed that over time, peers in Radical crowds socialized each other into becoming more socially anxious. More research is needed in order to probe the exact processes behind the connection between Radical crowd affiliation and seemingly poor internal adjustment. Perhaps creating tolerance, openness, and an inviting atmosphere in settings where crowds are most salient, such as schools, may help prevent or diminish some of these problems in the first place. References Bauman, K. E., & Ennett, S. T. (1994). Peer influence on adolescent drug use. American Psychologist, 49, 820–822. Bem, D. J. (1967). Self-perception: an alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74, 183–200. Benenson, J. F. (1990). Gender differences in social networks. The Journal of Early Adolescence, 10, 472–495. J Abnorm Child Psychol (2011) 39:1239–1249 Bešić, N., & Kerr, M. (2009). Punks, Goths, and other eye-catching peer crowds: do they fulfill a function for shy youths? Journal of Research on Adolescence, 19, 113–121. Brown, B. B. (1990). Peer groups and peer cultures. In S. S. Feldman & G. R. Elliott (Eds.), At the threshold: The developing adolescent (pp. 171–196). Cambridge: Harvard University Press. Brown, B. B., & Lohr, M. J. (1987). Peer-group affiliation and adolescent self-esteem: an integration of ego-identity and symbolic interaction theories. Journal of Personality and Social Psychology, 52, 47–55. Brown, B. B., Lohr, M. J., & Trujillo, C. (1990). Multiple crowds and multiple life styles: adolescents’ perceptions of peer-group stereotypes. In R. E. Muus (Ed.), Adolescent behavior and society: A book of readings, vol. 4 (pp. 30–36). NY: McGraw-Hill. Brown, B. B., Mory, M. S., & Kinney, D. (1994). Casting adolescent crowds in a relational perspective: Caricature, channel and context. In R. Montemayor, G. R. Adams, & T. P. Guillota (Eds.), Personal relationships during adolescence, vol. 6 (pp. 123–167). London: Sage Publications. Brown, B. B., Von Bank, H., & Steinberg, L. (2008). Smoke in the looking glass: effects of discordance between self- and peer rated crowd affiliation on adolescent anxiety, depression and selffeelings. Journal of Youth and Adolescence, 37, 1163–1177. Bruch, M. A., & Cheek, J. M. (1995). Developmental factors in childhood and adolescent shyness. In R. G. Heimberg, M. R. Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis, assessment, and treatment (pp. 163–182). NY: Guilford. Burk, W. J., Steglich, C. E. G., & Snijders, T. A. B. (2007). Beyond dyadic interdependence: actor-oriented models for co-evolving social networks and individual behaviors. International Journal of Behavioral Development, 31, 397–404. Byrne, D. (1971). The attraction paradigm. New York: Academic. Cairns, R. B., Leung, M.-C., Buchanan, L., & Cairns, B. D. (1995). Friendships and social networks in childhood and adolescence: fluidity, reliability, and interrelations. Child Development, 66, 1330–1345. Cohen, G. L., & Prinstein, M. J. (2006). Peer contagion of aggression and health risk behavior among adolescent males: an experimental investigation of effects on public conduct and private attitudes. Child Development, 77, 967–983. Cox, B. J., MacPherson, P. S. R., & Enns, M. W. (2005). Pscychiatric correlates of childhood shyness in a nationally representative sample. Behaviour Research and Therapy, 43, 1019–1027. Crozier, W. R. (1981). Shyness and self-esteem. British Journal of Social Psychology, 20, 220–222. Delsing, M. J. M. H., ter Bogt, T. F. M., Engels, R. C. M. E., & Meeus, W. H. J. (2007). Adolescents’ peer crowd identification in the Netherlands: structure and associations with problem behaviors. Journal of Research on Adolescence, 17, 467–480. Dishion, T. J., Spracklen, K. M., Andrews, D. W., & Patterson, G. R. (1996). Deviancy training in male adolescents friendships. Behavior Therapy, 27, 373–390. Elovainio, M., Kivimäki, M., Puttonen, S., Heponiemi, T., Pulkki, L., & Keltikangas-Järvinen, L. (2004). Temperament and depressive symptoms: a population-based longitudinal study on Cloninger’s psychobiological temperament model. Journal of Affective Disorders, 83, 227–232. Gren-Landell, M., Tillfors, M., Furmark, T., Bohlin, G., Andersson, G., & Svedin, C. G. (2009). Social phobia in Swedish adolescents. Social Psychiatry and Psychiatric Epidemiology, 44, 1–7. Haynie, D. L. (2001). Delinquent peers revisited: does network structure matter? American Journal of Sociology, 106, 1013–1057. Heaven, P. C. L., Ciarrochi, J., Vialle, W., & Cechavicuite, I. (2005). Adolescent peer crowd self-identification, attributional style and J Abnorm Child Psychol (2011) 39:1239–1249 perceptions of parenting. Journal of Community & Applied Social Psychology, 15, 313–318. Henrich, C. C., Kupermine, G. P., Sack, A., Blatt, S. J., & Leadbeater, B. J. (2000). Characteristics and homogeneity of early adolescent friendship groups: a comparison of male and female clique and nonclique members. Applied Developmental Science, 4, 15–26. Hogue, A., & Steinberg, L. (1995). Homophily of internalized distress in adolescent peer groups. Developmental Psychology, 31, 897–906. Huisman, M., & Snijders, T. A. B. (2003). Statistical analysis of longitudinal network data with changing composition. Sociological Methods Research, 32, 253–287. Huisman, M., & Steglich, C. (2008). Treatment of non-response in longitudinal network studies. Social Networks, 30, 297–308. Kandel, D. B. (1978). Homophily, selection, and socialization in adolescent friendships. The American Journal of Sociology, 84, 427–436. Kerr, M., Stattin, H., & Kiesner, J. (2007). Peers and problem behavior: Have we missed something? In R. C. M. E. Engels (Ed.), Friends, lovers and groups: Key relationships in adolescence. Hot topics in developmental research (pp. 125-153). New York: John Wiley & Sons. Kiesner, J., Cadinu, M., Poulin, F., & Bucci, M. (2002). Group identification in early adolescence: Its relation with peer adjustment and its moderator effect on peer influence. Child Development, 73, 196–208. Kinney, D. A. (1993). From Nerds to Normals: the recovery of identity among adolescents from middle school to high school. Sociology of Education, 66, 21–40. Kobus, K. (2003). Peers and adolescent smoking. Addiction Special Issue: Contexts and Adolescent Tobacco Use Trajectories, 98, 37–55. La Greca, A., & Lopez, N. (1998). Social anxiety among adolescents: linkages with peer relations and friendships. Journal of Abnormal Child Psychology, 26, 83–94. La Greca, A. M., Prinstein, M. J., & Fetter, M. D. (2001). Adolescent peer crowd affiliation: linkages with health-risk behaviors and close friendships. Journal of Pediatric Psychology, 26, 131–143. Laursen, B., Bukowski, W. M., Aunola, K., & Nurmi, J.-E. (2007). Friendship moderates prospective associations between social isolation and adjustment problems in young children. Child Development, 78, 1395–1404. Mackey, E. R., & La Greca, A. M. (2007). Adolescents’ eating, exercise, and weight control behaviors: does peer crowd affiliation play a role? Journal of Pediatric Psychology Special Issue: Pediatric Overweight, 32, 13–23. McCarroll, E. M., Lindsey, E. W., MacKinnon-Lewis, C., Campbell Chambers, J., & Frabutt, J. M. (2009). Health status and peer relationships in early adolescents: the role of peer contact, selfesteem, and social anxiety. Journal of Child & Family Studies, 18, 473–485. Newman, P. R., & Newman, B. M. (1976). Early adolescence and its conflict: group identity versus alienation. Adolescence, 11, 261–274. Pilkonis, P. A. (1977). The behavioral consequences of shyness. Journal of Personality, 45, 596–611. Prinstein, M. J. (2007). Moderators of peer contagion: a longitudinal examination of depression socialization between adolescents and their best friends. Journal of Clinical Child and Adolescent Psychology, 36, 159–170. Prinstein, M. J., & La Greca, A. M. (2002). Peer crowd affiliation and internalizing distress in childhood and adolescence: a longitudinal follow-back study. Journal of Research on Adolescence, 12, 325–351. Prinstein, M. J., & Wang, S. S. (2005). False consensus and adolescent peer contagion: examining discrepancies between perceptions 1249 and actual reported levels of friends’ deviant abd health risk behaviors. Journal of Abnormal Child Psychology, 33, 293–306. Radloff, L. S. (1977). The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. Romero, L. E., & Epkins, C. C. (2008). Girls’ cognitions of hypothetical friends: are they related to depression, loneliness, social anxiety and perceived similarity? Journal of Social and Personal Relationships, 25, 311–332. Rose, A. J. (2002). Co-rumination in the friendships of girls and boys. Child Development, 73, 1830–1843. Rosenberg, M. (1979). Conceiving the self. New York: Basic books. Santee, R. T., & Maslach, C. (1982). To agree or not to agree: personal dissent amid social pressure to conform. Journal of Personality and Social Psychology, 42, 690–700. Selfhout, M. H., Branje, S. J., & Meeus, W. H. (2009). Developmental trajectories of perceived friendship intimacy, constructive problem solving, and depression from early to late adolescence. Journal of Abnormal Child Psychology, 37, 251–264. Selfhout, M., Burk, W., Branje, S., Denissen, J., van Aken, M., & Meeus, W. (2010). Emerging late adolescent friendship networks and Big Five personality traits: a social network approach. Journal of Personality, 78, 509–538. Smith, H. M., & Betz, N. E. (2002). An examination of efficacy and esteem pathways to depression in young adulthood. Journal of Counseling Psychology, 49, 438–448. Snijders, T. A. B. (2001). The statistical evaluation of social network dynamics. Sociological methodology, 31, 361–395. Snijders, T. A. B., Steglich, C. E. G., Schweinberger, M., & Huisman, M. (2007). Manual for SIENA version 3. Groningen: University of Groningen. Snijders, T. A. B., van de Bunt, G. G., & Steglich, C. E. G. (2010). Introduction to actor-based models for network dynamics. Social Networks, 32, 44–60. Stednitz, J. N., & Epkins, C. C. (2006). Girls’ and mothers’ social anxiety, social skills, and loneliness: associations after accounting for depressive symptoms. Journal of Clinical Child and Adolescent Psychology, 35, 148–154. Stone, M., & Brown, B. B. (1998). In the eye of the beholder: adolescents’ perceptions of peer crowd stereotypes. In R. E. Muus (Ed.), Adolescent behavior and society: A book of readings (pp. 158–169). NY: McGraw-Hill. Sweden, S. (2004). Köping municipal facts. Retrieved from: http:// www.koping.se/kommunfakta/2004/Municipal_facts_2004.pdf. Thurlow, C. (2001). The usual suspects? A comparative investigation of crowds and social-type labelling among young British teenagers. Journal of Youth Studies, 4, 319–334. Urberg, K. A., Değirmencioğlu, S. M., Tolson, J. M., & HallidayScher, K. (2000). Adolescent social crowds: measurement and relationship to friends. Journal of Adolescent Research, 15, 427–445. Van Zalk, M. H. W., Kerr, M., Branje, S. J. T., Stattin, H., & Meeus, W. H. J. (2010). It takes three: selection, influence, and de-selection processes of depression in adolescent friendship networks. Developmental Psychology, 46, 927–938. Van Zalk, N., Van Zalk, M., Kerr, M., & Stattin, H. (2011). Social anxiety as a basis for friendship selection and socialization in adolescents’ social networks. Journal of Personality, 79, 499– 526. Wojslawowicz Bowker, J. C., Rubin, K. H., Burgess, K. B., Booth-LaForce, C., & Rose-Krasnor, L. (2006). Behavioral characteristics associated with stable and fluid best friendship patterns in middle childhood. Merrill-Palmer Quarterly, 52, 671–693.