Socialization of Physical and Social Aggression ... - Wiley Online Library

doi: 10.1111/j.1467-9507.2011.00621.x

Socialization of Physical and Social Aggression in Early Adolescents’ Peer Groups: High-status Peers, Individual Status, and Gender sode_621

170..194

Bing Shi, Sun Yat-sen University and Hongling Xie, Temple University

Abstract The influence of high-status peers on a target individual’s physical and manipulative social aggression in peer groups was examined in a diverse sample of seventh-grade students. A total of 245 individual members belonging to 65 groups were included in analyses. Aggression was assessed by peer and victim nominations in the fall and spring semesters of seventh grade. High-status peers rather than low-status peers in a group had a strong influence on individual members’ physical and social aggression. High-status peers were particularly influential on low-status individual members’ social aggression. A similar pattern was found for physical aggression in boys’ groups. These findings imply that high-status members’ aggression rather than the average of all members’ may better represent the group norm. Special attention needs to be given to high-status aggressive adolescents in future intervention and prevention of aggression in schools. Keywords: peer group; social status; physical aggression; social aggression

Introduction Developmental psychologists and sociologists have acknowledged the important role of peer groups in shaping and supporting group members’ behavior (e.g., Corsaro & Eder, 1990; Crockett, Losoff, & Peterson, 1984; Eder, 1985; Harris, 1995; Rubin, Bukowski, & Parker, 2006). Antisocial behavior, such as aggression, is found to increase over time for adolescents who affiliate with antisocial friends or peers in a

This research was supported by grants from the National Science Foundation (NSF 0339070) and William T. Grant Foundation (Grant ID 6934) to Hongling Xie. The views expressed in this paper are those of the authors’ and do not necessarily reflect the views of the granting agencies. We thank the children who participated in our study and the schools that assisted our study in various ways. We appreciate the assistance by Tabitha J. Wurster during manuscript preparation and the assistance by Pinky Patel, Ashley Dugan, Ngalula Fleurant, Marisa Gauger, Sara Heverly-Fitt, Phylicia Joseph, and Olivia Taduran in data collection and processing. Correspondence should be addressed to Bing Shi, Department of Marketing, Sun Yat-sen University, No. 135, West Xingang Road, Guangzhou, China 510275. Email: [email protected] © Blackwell Publishing Ltd. 2011. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

Socialization of Physical and Social Aggression in Peer Groups 171 group (e.g., Dishion & Owen, 2002; Ellis & Zarbatany, 2007; Espelage, Holt, & Henkel, 2003). However, previous studies failed to further consider potential variability in two important aspects: (1) the strength of influences of different peers in a social group; and (2) the degree to which each group member is influenced by peers. Variability in these two aspects has been considered in a conceptual model of peer influence proposed by Brown and colleagues (Brown, Bakken, Ameringer, & Mahon, 2008), in which several factors have been speculated to affect an individual’s strength of influence and openness to influence. Note that this model concerns peer influence in general. In this study, we focused on peer influence within early adolescents’ natural social groups in which frequent social interactions occur among group members. Several developmental and social psychology theories suggest that social status is a likely factor associated with variability in an individual’s strength of influence on peers and openness to influence by peers. For instance, social learning theory (Bandura, 1986), suggests that a peer’s social status would affect the strength of his or her influence on an individual. In addition, social impact theory (Latané, 1981), proposes that an individual’s social status would affect his or her openness toward peer influences. We selected early adolescence on the basis that peers play important roles in early adolescents’ development (Brown, 1990; Hartup, 1993), and that peer groups are the primary context for early adolescents’ interactions with peers (Crockett et al., 1984). Socialization in Peer Groups Several mechanisms of socialization within peer groups have been proposed. One mechanism focuses on social interactions that occur within a group. It is proposed that interactional synchrony and socialization could increase the similarities among group members (Cairns, 1979; Dishion, McCord, & Poulin, 1999; Patterson, DeBaryshe, & Ramsey, 1989). Individuals’ actions serve to constrain and/or reinforce others’ actions in a group. Over time, synchronized interactions and social reinforcement could enhance the levels of similarity among group members. Findings from multiple studies support this notion: Affiliation with aggressive or delinquent peers increases a child’s aggressive or problematic behaviors later in development (e.g., Cairns & Cairns, 1994; Crosnoe & Needham, 2004; Dishion & Owen, 2002; Ellis & Zarbatany, 2007; Espelage et al., 2003; Xie, Cairns, & Cairns, 2001). Another mechanism is based on social comparisons between individuals and other group members (Festinger, 1954). It is proposed that humans have a fundamental need to compare themselves with others to evaluate their own abilities and opinions. The existence of any discrepancy between an individual and other members in a group will lead to changes either in the individual or in other members so that the discrepancy will be reduced (Festinger, 1954). Findings from Boxer, Guerra, Huesmann, and Morales’ (2005) study are consistent with this proposition. They found that the level of aggression in a target early adolescent increased if his or her group mates were more aggressive than him or her, and decreased if group mates were less aggressive. A third mechanism involves normative influence processes (e.g., Kruglanski & Webster, 1991). It is proposed that an individual member is motivated to learn and adhere to the group norm in part because of concerns over experiencing negative consequences such as embarrassment, awkwardness, and even hostile rejection by others that may be caused by one’s deviation from the norm. Taken together, all the three mechanisms imply that members in a group would become similar to each other over time. © Blackwell Publishing Ltd. 2011

Social Development, 21, 1, 2012

172

Bing Shi and Hongling Xie

The Influence of High-status Peers in a Group Note that the mechanisms mentioned fail to consider different levels of influences by group members. In other words, it is not clear what kinds of members are likely to cause changes in others or to set the group norm. In early adolescence, one emerging characteristic of a peer group is its hierarchical structure (Harris, 1995). Members in a group often vary in social status (e.g., Adler & Adler, 1998; Dunphy, 1963; Hartup, 1993). Based on social impact theory (Latané, 1981), group members with various levels of social status would vary in their strength of social impact on a target individual. The theory postulates that high-status members would have a stronger social impact than low-status members on a target individual. Similarly, Brown and colleagues (2008) speculate that the strength of peer influence may be affected by the salience of the influencer which may be related to the social status of this influencer among other factors. However, neither the social impact theory nor the peer influence model developed by Brown et al. (2008) specify the processes of high-status peers’ influences. Similarly, social learning theory (Bandura, 1986), proposes that high-status peers would have a stronger influence than low-status peers on an individual. However, this theory further purports potential processes by which high-status peers would exert a stronger influence than low-status peers. It is proposed that children’s behavior is learned through observation of others’ behavior (Bandura & Walters, 1963). Observational learning is governed by four processes: attention; retention; production; and motivation. Peers with different levels of social status would affect the attention and motivation processes of children’s observational learning to different degrees. Compared with low status peers, high-status peers receive more peer attention and possess social power in the peer network (e.g., Vaughn & Langlois, 1983). The increased visibility of high-status peers means that they are more likely to receive attention from and to be modeled by an individual than low-status peers. In addition, since high levels of social status are associated with social and material benefits (e.g., Hawley, 1999), an individual would be motivated to imitate aggressive behaviors displayed by high-status peers in an effort to gain access to the benefits associated with high status. This motivation aspect is consistent with a goal-framing approach (Lindenberg, 2001, 2006). It is proposed that a target individual’s goal would affect all the processes involved in observational learning (e.g., attention, production). One important goal for early adolescents is to establish relations with peers and to strive for not only group membership but also for high social status in peer groups or networks (e.g., Baumeister & Leary, 1995; Buhrmester, 1990; LaFontana & Cillessen, 2010; Jarvinen & Nicholls, 1999; Ojanen, Grönroos, & Salmivalli, 2005). For example, findings in LaFontana and Cillessen’s (2010) study show that early adolescents prioritize the attainment of social status among peers over other domains such as friendships, athletic/academic achievements, and following school rules. Moreover, imitating high-status peers could be an effective means of maintaining group membership and enhancing one’s social status (Cialdini et al., 1976; Cialdini & Richardson, 1980; Dijkstra, Cillessen, Lindenberg, & Veenstra, 2010). All these theories strongly suggest that high-status peers would be more influential than low-status peers in an early adolescent. This prediction is supported by two recent studies (Cohen & Prinstein, 2006; Ellis & Zarbatany, 2007). In a computer-simulated ‘chat room’, Cohen and Prinstein found that peers with high levels of social status induced more conformity from an adolescent to their responses on aggressive and risk scenarios than peers with low levels of social © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 173 status did. Note that this study was conducted in a simulated lab setting, and peer influences in natural peer social networks were not examined. The other study focuses on the effect of social status of a peer group on aggression and deviant behavior (Ellis & Zarbatany, 2007). High-status of a group (i.e., high-group centrality or visibility) was found to magnify the socialization of aggression and deviant behavior. Social status in Ellis and Zarbatany’s study represents group status. To our knowledge, no previous studies have explicitly examined in a natural peer social group how members of different social statuses vary in their influences. The strong influence of high-status members in a group implies that they are likely to be the ones to set up the group norm. Crandall (1988) found that a college women’s binge eating was influenced by the average binge-eating level of her friendship group, and further proposed that the group average might not always represent the normative level of a group. Rather, popular or high-status women may set the tone for a group, and their binge-eating level may be transmitted to the rest of group members. Consistent with this speculation, a recent cross-sectional study (Dijkstra, Lindenberg, & Veenstra, 2008) shows that it is the level of bullying of popular children rather than that of non-popular children in a class that moderates the acceptance of bullying behavior by peers in early adolescence. The higher the level of bullying by popular youth, the less negative the association between bullying and peer acceptance in the class. Individual Status in a Group Previous studies show that not all children are equally influenced by their peers. One factor which may moderate the extent to which individuals are influenced by peers is the individual’s level of resistance or openness to peer influences (Berndt, 1979; Erickson, Crosnoe, & Dornbusch, 2000; Monahan, Steinberg, & Cauffman, 2009; Steinberg & Monahan, 2007). For instance, Monahan et al.’s (2009) study reports that peer influences on antisocial behaviors were stronger among youth who are low in resistance to peer influence than among youth who are high in resistance. An adolescent’s openness to peer influence may be affected by many factors. According to social impact theory, a person’s social status in a social unit would influence his or her openness toward the other members’ influence, particularly toward high-status members’ (Latané, 1981). A low-status person would be more open than a high-status person toward the impact of high-status others (Latané, 1981). In addition, Brown et al. (2008) propose that within a group, a low-status individual in a group may enhance his or her openness to peer influence in order to remain in the group. Both social impact theory and the proposed group dynamics predict that a low-status individual in a social group would be more influenced by high-status peers than a high-status individual would. We tested this prediction in this study. Note that rarely have previous developmental studies considered that group members may be influenced to different degrees. Measures of Social Status Three constructs of social status have been reported in previous research: peer preference, perceived popularity, and network centrality. Peer preference evaluates a child’s relative likeability by peers (e.g., Coie, Dodge, & Coppotelli, 1982; Lease, Musgrove, & Axelrod, 2002). It is measured by subtracting the number of times being nominated by peers as liked-least (an index of repulsion) from the number of times © Blackwell Publishing Ltd. 2011


174


being nominated as liked-most (an index of attraction). Perceived popularity and network centrality, on the other hand, highlight an individual’s prominence in peer networks. Perceived popularity is often assessed by peer nominations of the most popular children in school (e.g., Gorman, Kim, & Schimmelbusch, 2002; LaFontana & Cillessen, 2002; Prinstein & Cillessen, 2003). Network centrality was first advanced by Cairns and colleagues in the social cognitive map (SCM) procedure for identifying peer groups (e.g., Cairns, Gariépy, & Kindermann, 1991; Cairns, Leung, Buchanan, & Cairns, 1995; Cairns, Perrin, & Cairns, 1985; Cairns, Xie, & Leung, 1998). In this SCM procedure, network centrality is measured by the number of times a child is nominated by peers into a social group. High levels of perceived popularity and network centrality are often associated with being admired, influential, and a leader (e.g., Farmer et al., 2002; Lease et al., 2002; Rodkin, Farmer, Pearl, & Van Acker, 2000; Rubin & Coplan, 1992). In the perspectives of social learning theory and social impact theory, high-status individuals represent those who are likely to be imitated by others and have a strong power of social impact. Therefore, perceived popularity and network centrality rather than peer preference are likely to be associated with varying levels of peer influences. In this study, we selected social centrality over perceived popularity given that we aimed to investigate peer influence of aggression within a social group. An index of an individual’s status within a group would be more appropriate than measures of social status in the broad network. Perceived popularity measures an adolescent’s status in the whole social network rather than within a specific peer group. The mechanisms of peer influence in the broad social network may be quite different from those in peer groups (e.g., direct social interaction, normative process). Therefore, we selected the centrality measure produced by the SCM procedure which also yields critical information on individual status within a group and on group membership (e.g., Cairns, Cairns, Neckerman, Gest, & Gariépy, 1988; Cairns et al., 1985, Cairns et al., 1998). The Present Study Considering that aggression is a social behavior woven into the fabrics of social interactions (Cairns & Cairns, 1994), we aimed to investigate high-status peers’ influence on individuals’ aggression in peer groups. Various forms of aggression have been identified in research. Two forms of aggression, physical and social aggression, were examined. Previous studies have concentrated on the socialization of physical aggression in peer groups. Although subtle forms of aggression, such as social aggression which often refers to actions that cause interpersonal damage and are achieved by concealed methods that employ the social community (Cairns, Cairns, Neckerman, Ferguson, & Gariépy, 1989; Xie, Cairns, & Cairns, 2002), are frequently used and increasingly important in early adolescence (e.g., Cairns et al., 1989; Cillessen & Mayeux, 2004; Xie et al., 2002), rarely have studies investigated subtle forms of aggression in groups with the exception of one recent report (Ellis & Zarbatany, 2007). It should be noted that there are significant gender differences in physical aggression. Physical aggression is more common in boys than in girls (e.g., Archer, 2004; Card, Stucky, Sawalani, & Little, 2008; Maccoby & Jacklin, 1974). Some studies indicate that the association between physical aggression and social status varies by gender in early adolescence. A positive association between physical aggression and social status (i.e., network centrality and perceived popularity) is often found for boys but not always found for girls (e.g., for network centrality, Farmer & Rodkin, 1996; © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 175 Rodkin et al., 2000; Xie, Cairns, & Cairns, 1999; for perceived popularity, Cillessen & Mayeux, 2004; LaFontana & Cillessen, 2002; Rose, Swenson, & Waller, 2004). Based on gender differences mentioned above, it can be expected that the influence of high-status peers on individual members’ physical aggression may be stronger in boys’ groups than in girls’ groups.1 Moreover, potential moderating effect of individual status on peer influences would be more evident in boys’ groups than in girls’ groups. These expectations are supported by findings from previous studies indicating that in early adolescence boys are more susceptible or less resistant than girls to peer influences (e.g., Berndt, 1979; Brown, Clasen, & Eicher, 1986; Devereux, 1970; Steinberg & Monahan, 2007; Steinberg & Silverberg, 1986). In contrast, gender differences in social aggression are not consistent in previous research. Although some studies report that girls score higher in social/relational/ indirect aggression than boys (e.g., Crick & Grotpeter, 1995; Lagerspetz, Björkqvist, & Peltonen, 1988), others report no reliable gender differences or that boys score higher than girls (e.g., Salmivalli & Kaukiainen, 2004; Tomada & Schneider, 1997; Underwood, Scott, & Galperin, 2004). A meta-analytic review finds trivial gender difference in indirect aggression (Card et al., 2008). More importantly, the association between social aggression and social status is often similar across the genders. Recent reports show that in middle childhood and early adolescence, social aggression is associated with high levels of social status among both boys and girls (e.g., for network centrality, Xie, Swift, Cairns, & Cairns, 2002; for perceived popularity, LaFontana & Cillessen, 2002; Lease et al., 2002). Overall findings in previous studies suggest that peer influences on social aggression may be more similar in the two genders than different.

Method Participants and Procedure A total of 321 seventh-grade students (55 percent girls, mean age = 13.01, SD = 0.45, ranging from 12.19 to 14.25) participated in this study. Participants were recruited from three middle schools in an urban school district in the Northeastern USA. Only students with signed parental consent were included. The overall participation rate for this study was 59 percent (321/547), with separate participation rate for each of school as 58 percent (131/224), 57 percent (92/162), and 61 percent (98/161). A higher participation rate was found among girls (68 percent) than among boys (50 percent), c2 (1, 547) = 17.77, p < .001. About one half of participants were African-American, 31 percent White people, 19 percent Hispanic, and 1 percent Asian or other ethnicity. Fifty percent of participants were eligible for free lunch at school, and an additional 8 percent of participants were eligible for reduced-price lunch. Participants and nonparticipants did not differ in their ethnicity background and eligibility for free or reduced-price lunch, c2 s < 6.95, ps > .13. In the fall and spring semesters of seventh grade, participants completed questionnaire measures in a group administered survey session. The two data points were spaced 5 to 6 months apart (e.g., early December and mid May). During the survey, a lead administrator read the instructions and questions aloud, while additional assistants provided mobile monitoring and assistance to students. Before the beginning of the survey, participants were assured of confidentiality, and they were asked not to discuss their responses with other students. A blank paper was provided to each participant to © Blackwell Publishing Ltd. 2011


176


cover their answers in the survey. Participants were also told they could stop their participation at any time without any repercussion. Measures Peer Groups. In the fall semester of seventh grade (time 1), participants were asked to nominate peer groups in their grade: ‘Are there any students in your grade who hang around together a lot?’ ‘Please write their names.’ Following previous studies (e.g., Cairns et al., 1985, 1988), no name list was provided, and participants nominated groups from free recall. According to nominations of peer groups from all respondents in a grade, the SCM procedure was employed to identify peer groups (e.g., Cairns & Cairns, 1994; Cairns et al., 1991; Cairns et al., 1985). In previous studies, groups identified by the SCM procedure have been validated by direct observations of social interactional patterns (e.g., Cairns et al., 1985; Gest, Farmer, Cairns, & Xie, 2003). Group membership was identified for 272 participants, belonging to 69 peer groups (with three or more members). Group size ranged from 3 to 16 (M = 6.58, SD = 3.09). The majority of identified groups were of the same gender: 32 all-girl groups, 29 all-boy groups, and eight mixed-gender groups. All-girl groups did not differ from all-boy groups in group sizes, t (59) = 1.14, p = .26. In order to test differences across group gender, we excluded four groups with mixed-gender high-status members. In the other four mixed-gender groups, we excluded five low-status individuals whose gender was different from their high-status members’ gender. A total of 245 individual members belonging to 65 peer groups were included in the analyses. Participants included in the analyses did not differ from participants excluded from analyses and from non-participants in their ethnicity background and eligibility for free or reducedprice lunch, c2s < 1.36, ps > .24. High-status and Low-status Members in a Group. Information from the SCM analysis was used to identify high-status and low-status members in a group. According to the original SCM algorithm (e.g., Cairns & Cairns, 1994; Cairns et al., 1988), individual nomination was determined by the number of times an individual was nominated into a group. Group nomination was determined by the average of individual nominations for the two most nominated members in a group. An individual was classified as nuclear in a group if his or her nomination was greater than or equal to 70 percent of the group nomination (i.e., .7 * group nomination). An individual was classified as peripheral if his or her nomination was below 30 percent of the group nomination (i.e., .3* group nomination). The rest were classified as secondary members in a group. In this study, high-status members in a group referred to nuclear members (mean number = 2.25, SD = 1.27), and low-status members in a group included secondary and peripheral members (mean number = 2.64, SD = 2.07). Note that not every group contained low-status members. In 17 groups, all members were high-status members. In the rest 48 groups, 52 percent of group members (ranging from 18 percent to 83 percent) were high-status members. Individual Status in a Group. A continuous variable was used to measure individual status in a group. In this study, individual status within a group was measured by individual nomination, which was standardized within group to control for group differences in nominations.2 © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 177 Measures of Aggression. Both forms of aggression, physical and social aggression, were measured by peer nominations and victim nominations. Participants were asked to nominate, from free recall, unlimited number of peers in the same grade who best fit descriptors for physical aggression—‘starts fights’ (‘This person starts fights. This person pushes them, or hits them.’), and for manipulative social aggression—‘sets up fights’ (‘this person is good at causing people to get mad at each other.’). For each student, the total number of peer nominations was calculated and standardized by school. In addition, participants were asked to report their victimization experience at school. First, they were asked if they had experienced specific types of peer victimization. They were then asked to nominate peers who had done this to them. The number of times a student was nominated for ‘hit, pushed, or kicked’ on a victim was used to assess physical aggression. The number of times a student was nominated for ‘gotten friends to turn against’ a victim was used to assess manipulative social aggression. For each student, the total number of victim nominations was calculated and standardized by school. Peer and victim nominations require participants to nominate other children who are aggressive in school from complimentary perspectives. Victim nominations focus on identifying individuals who have aggressed against a participant himself or herself, whereas peer nominations focus on identifying aggressive individuals regardless of whom they aggressed against. Conceptually, these two measures provide complementary information. Empirically, the standardized scores of peer nominations were significantly correlated with the standardized scores of victim nominations, r = .33, p < .001 for physical aggression, and r = .31, p < .001 for social aggression. Based on both conceptual consideration and empirical evidence, we calculate an average score across peer nominations and victim nominations for each form of aggression. Group Members’, High-status Members’, and Low-status Members’ Aggression. An average score of aggression across group members at time 1 was calculated to measure group members’ aggression. Similarly, an average score of aggression across highstatus (i.e., nuclear) members and across low-status (i.e., secondary and peripheral) members at time 1 was calculated as high-status members’ and low-status members’ aggression, respectively. Four groups excluded from analyses due to mixed-gender high-status members did not differ from the rest 65 groups included in the analyses in group members’ and high-status members’ physical and social aggression, ts < .68, ps > .50. Stability of Peer Affiliation. Peer groups in childhood and adolescence are often dynamic in organization (Brown, 1990; Cairns et al., 1995). Group membership is not static over time. Since the strength of high-status members’ influence would vary by the amount of time an individual child spent with high-status members, we controlled the stability of affiliating with high-status members in our analyses involving highstatus members’ influences. Using the same SCM procedure, we identified peer groups in the spring semester of seventh grade (time 2). We determined a target individual’s stability of affiliation with high-status members by comparing high-status members in a target individual’s groups at both time points (i.e., fall and spring semesters). First, we counted the number of high-status members who remained in the same group with a target child (excluding the target child if he or she was a high-status member). Then we divided this number by the total number of high-status members in the target child’s group at time 1 (again excluding the target child if he or she was a high-status © Blackwell Publishing Ltd. 2011


178


member). Stability of affiliating with high-status peers (Mean = .49, SD = .44) was significantly correlated with individual status in a group, r = .19, p < .01. Individuals with high level of social status were more stable than individuals with low level of social status in their affiliation with high-status peers in a group. Similarly, we controlled for the stability of a child’s affiliation with group members and low-status members, when we examined the influences of group members and low-status members. We coded the stability of affiliation with group members (Mean = .45, SD = .37) by dividing the number of group members who remained in the same group with a target child across both time points by the total number of members in this child’s time 1 group (excluding the target child). The stability of affiliation with low-status members (mean = .25, SD = .34) was coded by dividing the number of low-status members who remained in the same group with a target child across both time points by the total number of low-status members in this child’s group at time 1 (excluding the target child if he or she had low status). The stability of affiliation with group members was higher among higher-status individuals, r = .16, p < .05. However, individual status was not associated with the stability of affiliation with low-status members, r = .10, p = .14. Overview of Statistical Analyses Hierarchical linear modeling (HLM) procedure was used to test hypotheses in this study. The use of HLM procedure is necessary due to the hierarchical structure of data in this study: Individual members were nested within peer groups (Luke, 2004; Raudenbush & Bryk, 2002).We constructed two-level HLM models to test high-status members’ influence on a target individual member’s aggression at time 2 (spring semester) and to examine the moderating effects of individual status and group gender on high-status members’ influence. An individual’s aggression at time 1 (fall semester) and the stability of peer affiliation from time 1 to time 2 were entered in all level 1 models as control variables. For both physical and social aggression, two sets of analyses were conducted. In the first set of analyses, we compared the influences of all group members, high-status members, and low-status members on an individual member’s aggression over time. At level 1 (or individual level), the variance in individual members’ time 2 aggression was modeled. At level 2 (or group level), the randomly varying intercept from the level 1 model was entered as the dependent variable and peers’ aggression was entered as a predictor. For example, the level 1 model for testing high-status members’ influence is represented by the following equation: Individual aggression at time 2 = Β0j + Β1 j (individual aggression at time 1) + Β2 j (stability of affiliating with high-status members in a group) + rij (individual error)

The level 2 model for testing high-status members’ socialization effect is represented by the following equation: B0 j = γ 00 + γ 01 (high-status members’ aggression at time 1) + μ 0 j (group error)

The significance of the coefficient of high-status members’ aggression at time 1 (g01) determined whether high-status members had a significant influence on an individual’s aggression. Similar models were constructed to test the influences of group norm and © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 179 low-status peers. In the level 1 models, the stability of affiliating with other group members or low-status peers was controlled, and in the level 2 models, all group members’ aggression or low-status members’ aggression was entered. The second set of HLM analyses tested the moderating effects of individual status and group gender on high-status members’ influence. In this set of analyses, stability of affiliating with high-status members in a group was entered in all level 1 models as a control variable as well as individual aggression at time 1. To examine the moderation by individual status, we used the following level 1 model: Individual aggression at time 2 = Β0 j + Β1 j (individual status) + Β2 j (individual aggression at time 1) + Β3 j (stability of affiliating with high-status members) + rij (individual error)

The level 2 model was represented by the following equations: Β0 j = γ 00 + γ 01 (high-status members’ aggression at time 1) + μ 0 j (group error) Β1 j = γ 10 + γ 11 (high-status members’ aggression at time 1) + μ1 j (group error)

The significance of the coefficient of high-status members’ aggression (g11) on the varying slope of individual status determined whether individual status significantly moderated high-status members’ influence. To test the moderating effect of group gender, we used the same level 1 model for testing the influence of high-status members in the first set of data analyses. The level 2 model is described below: Β0 j = γ 00 + γ 01 (high-status members’ aggression at time 1) + γ 02 (group gender) + γ 03 (group gender × high-status members’ aggression at time 1) + μ 0 j (group error)

The significance of the coefficient of group gender x high-status members’ aggression at Time 1 (g03) determined whether group gender significantly moderated highstatus members’ influences. Most of the predictors in the level 1 models and all predictors in the level 2 models were grand-mean centered. Individual status in a group was group-mean centered so that the mean individual status would be the same across groups. Group gender was not centered, because it was a dichotomous variable. When using grand-mean centering, the level 2 model represents the group-level relationship between the predictor and the outcome adjusted for the influence of the level 1 predictor (Bryk & Raudenbush, 1992). Significant interactions in HLM models were presented according to the guideline outlined by Aiken and West (1991). Variables in the interaction were split into 1 SD above the mean (high level) and 1 SD below the mean (low level) to plot the effect. Results Correlations among Variables Table 1 lists the means and standard deviations of individual-level variables along with zero-order correlations among them. Children’s physical aggression was positively correlated with their social aggression, rs > .33, ps < .001. Boys tended to be more physically aggressive than girls at time 1, r = .11, p = .08; this trend was not found at time 2. No gender difference was found in social aggression at both time points. © Blackwell Publishing Ltd. 2011


180


Table 1. Means, Standard Deviations, and Correlations for Individual-Level Variables Variable 1 Gender (female coded as 0) 2 Individual status in a group 3 PHY at time 1 4 PHY at time 2 5 SOC at time 1 6 SOC at time 2

M (SD)

2

3

—

.01

.23 (.91) .09 .02 .16 .10

(.91) (.66) (1.00) (.82)

4

5

6

.11†

.02

-.09

-.09

—

-.004

-.01

.03

.03

— — — —

— — — —

.54*** .34*** — —

.38*** .42*** .56*** —

.35*** — — —

Notes: PHY = physical aggression; SOC = social aggression. † p = .08. *** p < .001.

Table 2 lists the means and standard deviations of group-level variables along with zero-order correlations among them. High-status members in boys’ groups were more physically aggressive than high-status members in girls’ group, r = .14, p < .05 whereas low-status members did not differ across group genders, r = .11, p = .11. In contrast, no difference in high-status members’ social aggression was found across boys’ and girls’ groups, r = .02, p = .81 whereas low-status members in girls’ groups were more socially aggressive than low-status members in boys’ groups, r = -.30, p < .001. Physical Aggression High-status Members’ Influence. We first tested the influence of group members, high-status members, and low-status members on an individual’s physical aggression. An individual’s physical aggression at time 1 and the stability of peer affiliation were entered as control variables. Results from HLM analyses are presented in Table 3. The coefficient for the average of group members’ aggression was significant, b = .25, t (60) = 2.48, p < .05. The higher the average levels of all members’ aggression, the more aggressive an individual member became over time. The coefficient for highstatus members’ aggression was significant, b = .16, t (60) = 2.55, p < .05. The more aggressive high-status members were, the more aggressive individual members became at time 2. But the coefficient for low-status members’ aggression was not significant, b = .11, t (46) = 1.32, p = .19. The aggression levels of low-status members did not influence an individual’s physical aggression over time. When we simultaneously examined the influences of high-status and low-status members on an individual’s physical aggression, only the coefficient for high-status members’ aggression was significant, b = .14, t (45) = 1.37, p = .05. The coefficient for low-status members’ aggression was not significant, b = .08, t (45) = .95, p = .35. These results indicate that it was the aggression of high-status members rather than that of low-status members that significantly influenced an individual’s physical aggression. Moderating Effects of Individual Status and Group Gender. Next, we tested the moderating effects of individual status and group gender on high-status members’ © Blackwell Publishing Ltd. 2011


© Blackwell Publishing Ltd. 2011

.10 (.47) .13 (.67) .11 (.63) .15 (.45) .20 (.72) .12 (.65)

1 2 3 4 5 6 7

Notes: PHY = physical aggression; SOC = social aggression. † p < .10. * p < .05. ** p < .01. *** p < .001.

Group gender (female group coded as 0) Group members’ PHY High-status members’ PHY Low-status members’ PHY Group members’ SOC High-status members’ SOC Low-status members’ SOC

M (SD)

Variable .19** — — — — — —

2 .14* .85*** — — — — —

3 .11 .81*** .39** — — — —

4

Table 2. Means, Standard Deviations, and Correlations for Group-Level Variables at Time 1 6 .02 .47*** .60*** .25† .74*** — —

5 -.17** .49*** .43*** .59*** — — —

-.30*** .50*** .20 .70*** .73*** .01 —

7

Socialization of Physical and Social Aggression in Peer Groups 181


182


Table 3. Influences on Individuals’ Physical Aggression: The Effects of Group Members, High-status Members and Low-status Members Variables

Model 1

Model 2

Model 3

Model 4

Level-1 intercept Intercept Group members’ PHY High-status members’ PHY Low-status members’ PHY Control variables Individual Time 1 PHY Stability of affiliation

.01 (.04) .25* (.10) — — — .15† (.07) .05a (.10)

.002 (.04) — .16* (.06) — — .16* (.07) -.001b (.09)

.03 (.05) — — .11 (.08) — .19* (.08) .01c (.13)

.02 (.05) — .14* (.07) .08 (.08) — .17* (.08) -.01a (.12)

Notes: PHY = physical aggression. Standard errors are presented in the parentheses. a Stability of affiliation with group members. b Stability of affiliation with high-status members. c Stability of affiliation with low-status members. Model 1 and 2 were constructed for all 65 groups. Model 3 and 4 were constructed for 48 groups with high- and low-status members. † p = .06. * p < .05.

influences. Results are summarized in Table 4. The coefficient for the interaction between individual status and high-status members’ aggression was not significant, b = .0003, t (58) = .01, p = .996. Individual status did not moderate high-status members’ influences on individuals’ physical aggression. However, the coefficient for the interaction between group gender and high-status members’ aggression was significant, b = .27, t (58) = 2.11, p < .05. Group gender significantly moderated high-status members’ influences on physical aggression. In boys’ groups, individuals conformed to high-status boys’ physical aggression over time whereas in girls’ groups, individuals were relatively unaffected by the high-status girls’ physical aggression (Figure 1). Moreover, we tested the three-way interaction between individual status, group gender, and high-status members’ aggression. The coefficient for the three-way interaction approached significance, b = -.23, t (56) = -1.80, p = .08. This result suggested that the moderating effect of individual status on high-status members’ influences varied across boys’ and girls’ groups. In boys’ groups, aggressive high-status members induced a larger increase in physical aggression among low-status individuals than among high-status individuals (Figure 2, left panel). In girls’ groups, aggressive highstatus members induced a larger increase among high-status individuals than among low-status ones (Figure 2, right panel). Social Aggression High-status Members’ Influence. Results from HLM analyses regarding the influence of group members’, high-status members’, and low-status members’ manipulative social aggression on a target individual are presented in Table 5. Individuals’ social aggression at time 1 and stability of peer affiliation were entered as control variables. The coefficient for the average aggression of group members approached significance, b = .16, t (60) = 1.86, p = .07. The higher the average levels of all group members’ aggression, the more aggressive individual members tended to become over time. © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 183 Table 4. Moderating Effects of Individual Status and Group Gender on Highstatus Members’ Influences on Physical Aggression Variables

Model 1

Level-1 intercept Intercept .002 (.04) High-status members’ .16** (.06) PHY Group gender — Group gender ¥ — high-status members’ PHY Level-1 individual status in a group Intercept — High-status members’ — PHY Group gender — Group gender ¥ — high-status members’ PHY Control variables — Individual Time 1 .16* (.07) PHY Stability of affiliation -.001 (.09) with high-status members

Model 2

Model 3

Model 4

.001 (.04) .17* (.07)

-.003 (.05) .01 (.09)

-.01 (.06) .03 (.10)

— —

.03 (.08) .27* (.13)

.03 (.08) .26* (.13)

.0004 (.05) .0003 (.07)

— —

-.02 (.06) .12 (.09)

— —

— —

.04 (.09) -.23† (.13)

— .15† (.08)

— .16* (.07)

— .15† (.08)

-.001 (.09)

-.02 (.09)

-.03 (.10)

Notes: PHY = physical aggression. Standard errors are in the parentheses. † p < .10. * p < .05. ** p < .01.

Moreover, the coefficient for high-status members’ aggression was significant, b = .07, t (60) = 2.42, p < .05. The more socially aggressive high-status members were, the more socially aggressive individuals became later on. However, the coefficient for low-status members’ aggression was not significant, b = .04, t (46) = .59, p = .56. The aggression levels of low-status members did not influence individuals’ social aggression. When the effects of high-status and low-status members were tested simultaneously, only the coefficient for high-status members’ aggression was significant, b = .07, t (45) = 2.05, p < .05. The coefficient for low-status members’ aggression was not significant, b = .04, t (45) = .53, p = .60. These results indicated that high-status members rather than low-status members in a group had significant influences on individual group member’s social aggression. Moderating Effects of Individual Status and Group Gender. Next, we tested the moderating effects of individual status and group gender on high-status members’ influences on manipulative social aggression. Results are presented in Table 6. The coefficient for the interaction between individual status and high-status members’ aggression was significant, b = -.09, t (58) = -2.29, p < .05. Individual status © Blackwell Publishing Ltd. 2011


184

Bing Shi and Hongling Xie .25

PHY level for individuals at time 2

.2 .15 .1 .05

Girls' groups 0

Boys' groups

- .05 - .1 - .15 - .2 Low

High

PHY level for high-status members in a group at time 1

.3

.3

.2

.2

.1

0

Low-status individuals High-status individuals - .1

- .2

- .3



Figure 1. The Influences of High-status Members on Individuals’ Time 2 Physical Aggression in Girls’ Groups and Boys’ Groups.

.1

0

Low-status individuals High-status individuals

- .1

- .2

Low

High

PHY level for high-status members in a boys' group at time 1

- .3

Low

High

PHY level for high-status members in a girls' group at time 1

Figure 2. The Influences of High-status Members on Individuals’ Time 2 Physical Aggression by Group Gender (left—boys’ groups; right—girls’ groups) and Individual Status: High-status Individuals (1 SD above Mean) vs. Low-status Individuals (1 SD below Mean). significantly moderated high-status members’ influences on social aggression. As Figure 3 indicated, low-status individuals conformed to their high-status members’ social aggression more than high-status individuals did. The coefficient for the interaction between group gender and high-status members’ aggression was not significant, © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 185 Table 5. Influences on Individuals’ Social Aggression: The Effects of Group Members, High-status Members, and Low-status Members Variables Level-1 intercept Intercept Group members’ SOC High-status members’ SOC Low-status members’ SOC Control variables Individual time 1 SOC Stability of affiliation

Model 1

.11* .16†

(.04) (.08)

—

Model 2

.10*

Model 3

Model 4

(.04)

.15** (.05) —

.14** (.05) —

(.03)

—

— .07*

—

—

— .41*** (.09)

— .41*** (.08)

-.01a

(.12)

.03b

.04

(.10)

(.06)

— .48*** (.11) -.05c

(.12)

.07*

(.03)

.04

(.07)

— .42*** (.10) .01a

(.15)

Notes. Standard errors are presented in the parentheses. SOC = social aggression. a Stability of affiliation with group members. b Stability of affiliation with high-status members. c Stability of affiliation with low-status members. Models 1 and 2 were constructed for all 65 groups. Models 3 and 4 were constructed for 48 groups with high- and low-status members. † p < .10. * p < .05. ** p < .01. *** p < .001.

b = -.12, t (58) = -1.51, p = .14. Group gender did not moderate high-status members’ influences on individuals’ social aggression. Finally, the coefficient for the three-way interaction between individual status, group gender, and high-status members’ aggression was not significant, b = -.04, t (56) = -.34, p = .74. The moderation of individual status on high-status members’ influences did not vary across group gender. Discussion We found that in peer social groups, high-status peers rather than low-status peers influenced the development of an early adolescent’s aggressive behaviors. This was true for both physical and manipulative social aggression. Furthermore, the influences of high-status peers were stronger on low-status individuals than on high-status individuals. This pattern was found for social aggression in both boys’ and girls’ groups and for physical aggression in boys’ groups. To our knowledge, these findings were the first in developmental literature to demonstrate that during early adolescence, peers in a group differ in their strength of influence on individual members’ aggression, and that individual members also differ in the degrees to which they are influenced by peers. In both cases, variations can be predicted by social status. These findings are highly consistent with the integrated framework of socialization in peer groups that we outlined in the introduction, drawing from key theories in developmental and social psychology and from conceptual models of peer influence in general. © Blackwell Publishing Ltd. 2011


186


Table 6. Moderating Effects of Individual Status and Group Gender on Highstatus Members’ Influences on Social Aggression Variables Level 1 intercept Intercept High-status members’ SOC Group gender Group gender ¥ high-status members’ SOC Level 1 individual status in a group Intercept High-status members’ SOC Group gender Group gender ¥ high-status members’ SOC Control variables Individual time 1 SOC Stability of affiliation with high-status members

Model 1

.10* .07*

(.04) (.03)

Model 2

Model 3

Model 4

.11** (.04) .05 (.04)

.12** (.04) .17* (.08)

.12** (.04) .15* (.08)

— —

— —

—

—

— —

.005 -.09*

-.08 -.12

(.08) (.08)

-.07 -.13

— (.04) (.04)

(.08) (.08)

—

— —

-.08 -.06

(.07) (.10)

.18* -.04

(.08) (.10)

— —

— —

— —

— .41*** (.08)

— .43*** (.08)

— .41*** (.08)

— .43*** (.08)

.03

.06

.03

.06

(.10)

(.11)

(.12)

(.12)

Notes: Standard errors are presented in the parentheses. SOC = social aggression. * p < .05. ** p < .01. *** p < .001.

As anticipated, high-status peers rather than low-status peers in a group exert significant influences on a target individual’s aggression. This was true regardless of the form of aggression. These findings were consistent with extant theories such as social learning theory (Bandura, 1986; Bandura & Walters, 1963), and social impact theory (Latané, 1981), and with conceptual models of peer influence (e.g., Brown et al., 2008), which propose that high-status children are more likely than low-status children to have strong influences on others. Along with the proposition in Crandall’s (1988) study and findings in Dijkstra et al.’s (2008) study, our findings indicate that the significant effects of group average are mostly accounted for by high-status members’ aggression. They suggest that a group’s norm is likely to be set by its high-status members. Therefore, future studies should consider using high-status members’ aggression level rather than the average level of all group members, to measure group norm. Given the prominent influence of high-status peers, special attention should be given to high-status aggressive students in future intervention and prevention efforts in reducing aggression and violence in school. © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 187

SOC level for individuals at time 2

.25

.2

.15

.1

Low status individuals .05

High status individuals

0 Low

High

SOC level for high-status members in a group at time 1 Figure 3. The Influences of High-status Members on Individuals’ Time 2 Social Aggression by Individual Status: High-status Individuals (1 SD above Mean) vs. Low-status Individuals (1 SD below Mean). In addition, individual members of a group differed in their openness to high-status peers’ influence. Low-status individuals, as compared with high-status individuals, were more strongly influenced by high-status peers. This pattern was found for social aggression in both boys’ and girls’ groups and for physical aggression in boys’ groups. These findings are consistent with social impact theory (Latané, 1981), and the conceptual model outlined by Brown et al. (2008), which propose that low-status individuals would be more open to the others’ social impact than high-status individuals. They also suggest that once aggressive behaviors of high-status individuals are effectively prevented or controlled in peer social groups, such effects are likely to be carried over to low-status children in the same group. Note that findings from this research did not show exactly how the influence of high-status peers occurs in a peer group. An individual could be either actively or passively influenced by their high-status peers. Two of the three mechanisms reviewed in the introduction (i.e., social comparison and normative process) may help explain the process by which an individual actively conform to peers’ influence. For instance, an individual may intentionally compare his or her own aggressive behaviors with those of the high-status peers, and if any discrepancy exists, the individual may change or modify his or her aggressive behaviors to reduce this discrepancy (Festinger, 1954). Also, it is likely that high-status peers in a group may have the power and ability to set group norms, and individuals may be motivated to conform to such group norms in © Blackwell Publishing Ltd. 2011


188


order to remain in the group. Furthermore, social learning theory proposes that individuals, particularly low-status individuals, may actively observe and imitate highstatus peers’ aggressive behaviors, given the prominence and benefits associated with high levels of social status (e.g., Bandura & Walters, 1963; Dijkstra et al., 2010; Fiske, 1993). Individuals could also be passive recipients of influences. From the perspective of social synchrony and deviancy training during social interactions (Cairns, 1979; Dishion, Sparcklen, Andrews, & Patterson, 1996), an individual’s aggressive behaviors may be positively reinforced by aggressive high-status peers. Such social reinforcement is likely to increase an individual’s level of aggression over time. In addition, individual members may increase their levels of aggression through peer coercion. For instance, aggressive high-status peers may victimize other members in the same group, particularly the low-status members, and thereafter these members may be coerced to fight back and become more aggressive over time. Future studies need to further identify the specific social processes involved in the peer influences within adolescent groups. In this study, we found that group gender moderated high-status peers’ influences for physical aggression but not for manipulative social aggression. High-status peers had a stronger influence on individual members’ physical aggression in boys’ groups than in girls’ groups. But no group gender difference was found for social aggression. These findings are consistent with our earlier speculations. Physical aggression is often more prevalent and more consistently associated with high status in peer networks among boys than among girls (e.g., Card et al., 2008; Farmer & Rodkin, 1996; Xie et al., 1999). As a result, the socialization of physical aggression could be stronger in boys’ groups than in girls’ groups. In contrast, gender differences in social aggression are less clear. As expected, no difference in group gender was found for social aggression. It should be noted that previous studies have rarely considered gender differences in the role of an individual’s status in peer influences. In our study, when we simultaneously investigated high-status peers’ influences and an individual’s status for each gender, an unexpected finding emerged for physical aggression in girls’ groups: With increasing levels of physical aggression among high-status girls, high-status female members in the same group increased their physical aggression more than low-status female members. In fact, the changes were in opposite directions for the high-status members and for the low-status members. As a result, the combined effect of high-status peers’ influences for the whole group was very weak for females. This explained the group gender differences in high-status peers’ influences for physical aggression. However, this unexpected finding on physical aggression in girls’ groups contradicts the pattern in boys’ groups and the predictions made by social impact theory (Latané, 1981). It is not clear why this pattern occurred in girls’ groups, and this finding itself needs to be replicated in future studies. We speculate that since physical aggression is less prevalent and accepted in girls than in boys, it may be that only high-status girls have the leverage to increase their levels of aggression. A second possibility could be the process of interactional synchrony through complementarity (e.g., Farmer, Xie, Cairns, & Hutchins, 2007). High-status girls may display physical aggression as a way to show or maintain their high levels of social status. Low-status girls, on the other hand, may support high-status girls’ physical aggression by complementing rather than by conforming to such behaviors. © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 189 Limitations and Future Directions Several limitations need to be discussed. A major limitation of this study is that only one group is allowed for each child as required by the nested relationship between an individual and a group in the HLM procedure. Around 20 percent of group members (55/272) in this study belonged to two groups and 3 percent (7/272) belonged to three groups. Because the requirements of multilevel analysis demanded that each child should be placed into only one peer group, those participants who belonged to multiple groups were assigned to their primary group where they received the most nominations. Children who affiliate with two or more peer groups may experience different types of influence. Future research comparing socialization effects across children’s multiple groups would help clarify this issue. Secondly, this study focused on socialization of aggression in groups. Participants who were isolates (18 percent, 60/336) or involved in dyads (1 percent, 4/336) were not included in the analyses. It is not clear how peers influence the development of these children’s aggression over time. Future research in this domain would provide us with more comprehensive information about the processes of peer influences for children not involved in peer groups. Thirdly, this study only tested the proposed conceptual model with children’s aggressive behaviors. It is not clear whether the model would be applied to other behavioral domains such as prosocial behaviors. Based on extant research indicating that prosocial behaviors could promote an individual’s social status in the network (e.g., Hawley, Little, & Card, 2008; Lease et al., 2002), similar processes of peer influences may be expected for prosocial behaviors. For other behaviors that are not associated with prominent social status in peer networks (e.g., academic competence in North American schools, see Becker & Luthar, 2007; Juvonen & Murdock, 1995; Wentzel & Asher, 1995), the proposed model of peer influences within social groups may need to be modified. Fourthly, this study focused on early adolescence, an important period when peer influences are strong. The model proposed in this study will need to be tested in children of different periods to investigate possible developmental differences. Finally, many factors may affect the variability in strength of a peer’s influence and the openness of an individual toward peer influence. In this study, we only investigated social status and demonstrated how social status could help understand the variability in both aspects. According to Brown et al.’s (2008) conceptual model, other factors such as the relation between an influencer and a recipient and group dynamics could also affect the process of peer influence. In the future, additional empirical studies are needed to further clarify the complex processes of peer influences in adolescent social groups. Summary Guided by a model integrating important developmental and social psychology theories (e.g., Bandura, 1986; Latané, 1981), and recent conceptualization of peer influences (Brown et al., 2008), this study is among the first to show that during early adolescence, peers in a social group differ in the strength of their influences on individual members’ aggression, and individual members also differ in their openness to peer influences. Variations in both aspects could be predicted by levels of social status. Specifically, high-status rather than low-status peers in a group were found to influence an early adolescent’s aggression. This was true regardless of the form of aggression. In addition, high-status peers had a stronger influence on low-status individual members than on high-status individual members. This pattern was found for © Blackwell Publishing Ltd. 2011


190


manipulative social aggression in boys’ and girls’ groups and also found for physical aggression in boys’ groups. These findings suggest that the average of aggression of high-status members rather than that of all members in a group should be measured as the group norm of aggression. Special attention should be given to high-status aggressive early adolescents in future interventions and preventions of aggression in school. References Adler, P. A., & Adler, P. (1998). Peer power: Preadolescent culture and identity. New Brunswick, NJ: Rutgers University Press. Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Archer, J. (2004). Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology, 8, 291–322. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A., & Walters, R. H. (1963). Social learning and personality development. New York: Holt, Rinehart and Winston. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. Becker, B. E., & Luthar, S. S. (2007). Peer perceived admiration and social preference: Contextual correlates of positive peer regard among suburban and urban adolescents. Journal of Research on Adolescence, 17, 117–144. Berndt, T. J. (1979). Developmental changes in conformity to peers and parents. Developmental Psychology, 15, 608–616. Boxer, P., Guerra, N. G., Huesmann, L. R., & Morales, J. (2005). Proximal peer-level effects of a small-group selected prevention on aggression in elementary school children: An investigation of the peer contagion hypothesis. Journal of Abnormal Child Psychology, 33, 325– 338. 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, MA: Harvard University Press. Brown, B. B., Bakken, J. P., Ameringer, S. W., & Mahon, S. D. (2008). A comprehensive conceptualization of the peer influence process in adolescence. In M. J. Prinstein, & K. A. Dodge (Eds.), Understanding peer influence in children and adolescents (pp. 17–44). New York: Guilford Press. Brown, B. B., Clasen, D. R., & Eicher, S. A. (1986). Perceptions of peer pressure, peer conformity dispositions, and self-reported behavior among adolescents. Developmental Psychology, 22, 521–530. Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models. Newbury Park, CA: Sage. Buhrmester, D. (1990). Intimacy of friendship, interpersonal competence, and adjustment during preadolescence and adolescence. Child Development, 61, 1101–1111. Cairns, R. B. (1979). Social development: The origins and plasticity of interchanges. San Francisco, CA: Freeman. Cairns, R. B., & Cairns, B. D. (1994). Lifelines and risks: Passages of youth in our time. New York: Cambridge University Press. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Ferguson, L. L., & Gariépy, J. L. (1989). Growth and aggression: I. Childhood to early adolescence. Developmental Psychology, 25, 320–330. Cairns, R. B., Cairns, B. D., Neckerman, H. J., Gest, S. J., & Gariépy, J. L. (1988). Social networks and aggressive behavior: Peer support or peer rejection? Developmental Psychology, 24, 815–823. Cairns, R. B., Gariépy, J. L., & Kindermann, T. (1991). Identifying social clusters in natural settings. Unpublished manuscript, University of North Carolina at Chapel Hill. 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. © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 191 Cairns, R. B., Perrin, J. E., & Cairns, B. D. (1985). Social structure and social cognition in early adolescence: Affiliative patterns. Journal of Early Adolescence, 5, 339–355. Cairns, R., Xie, H., & Leung, M. C. (1998). The popularity of friendship and the neglect of social networks: Toward a new balance. In W. M. Bukowski, & A. H. N. Cillessen (Eds.), Sociometry then and now: Building on six decades of measuring children’s experiences with the peer group (pp. 25–53). San Francisco, CA: Jossey-Bass Publishers. Card, N. A., Stucky, B. D., Sawalani, G. M., & Little, T. D. (2008). Direct and indirect aggression during childhood and adolescence: A meta-analytic review of gender differences, intercorrelations and relations to maladjustment. Child Development, 79, 1185–1229. Cialdini, R. B., Borden, R. J., Thorne, A., Walker, M. R., Freeman, S., & Sloan, L. R. (1976). Basking in reflected glory: Three (football) field studies. Journal of Personality and Social Psychology, 34, 366–375. Cialdini, R. B., & Richardson, K. D. (1980). Two indirect tactics of impression management: Basking and blasting. Journal of Personality and Social Psychology, 39, 406–416. Cillessen, A. H. N., & Mayeux, L. (2004). From censure to reinforcement: Development changes in the association between aggression and social status. Child Development, 75, 147–163. 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. Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982). Dimensions and types of social status: A cross-age perspective. Developmental Psychology, 18, 557–570. Corsaro, W. A., & Eder, D. (1990). Children’s peer cultures. Annual Review of Sociology, 16, 197- 220. Crandall, C. S. (1988). Social contagion of binger eating. Journal of Personality and Social Psychology, 55, 588–598. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710–722. Crockett, L. J., Losoff, M., & Peterson, A. C. (1984). Perceptions of the peer group and friendship in early adolescence. Journal of Early Adolescence, 4, 155–181. Crosnoe, R., & Needham, B. (2004). Holism, contextual variability, and the study of friendships in adolescent development. Child Development, 75, 264–279. Devereux, E. (1970). The role of peer group experience in moral development. In J. Hill (Ed.), Minnesota Symposium on Child Psychology (Vol. 4, pp. 94–140). Minneapolis, MN: University of Minnesota Press. Dijkstra, J. K., Cillessen, A. H. N., Lindenberg, S., & Veenstra, R. (2010). Basking in reflected glory and its limits: Why adolescents hang out with popular peers. Journal of Research on Adolescence, 20, 942–958. Dijkstra, J. K., Lindenberg, S., & Veenstra, R. (2008). Beyond the class norm: Bullying behavior of popular adolescents and its relation to peer acceptance and rejection. Journal of Abnormal Child Psychology, 36, 1289–1299. Dishion, T. J., McCord, J., & Poulin, F. (1999). When interventions harm: Peer group sand problem behavior. American Psychologist, 54, 755–764. Dishion, T. J., & Owen, L. D. (2002). A longitudinal analysis of friendships and substance use: Bidirectional influence from adolescence to adulthood. Developmental Psychology, 38, 480– 491. Dishion, T. J., Sparcklen, K. M., Andrews, D. W., & Patterson, G. R. (1996). Deviancy training in male adolescents friendships. Behavior Therapy, 27, 373–390. Dunphy, D. C. (1963). The social structure of urban adolescent peer groups. Sociometry, 26, 230–246. Eder, D. (1985). The cycle of popularity: Interpersonal relations among female adolescents. Sociology of Education, 58, 154–165. Ellis, W. E., & Zarbatany, L. (2007). Peer group status as a moderator of group influence on children’s deviant, aggressive, and prosocial behavior. Child Development, 78, 1240– 1254. Erickson, K. G., Crosnoe, R., & Dornbusch, S. M. (2000). A social process model of adolescent deviance: Combining social control and differential association perspectives. Journal of Youth and Adolescence, 29, 395–425. © Blackwell Publishing Ltd. 2011


192


Espelage, D. L., Holt, M. K., & Henkel, R. R. (2003). Examination of peer-group contextual effects on aggression during early adolescence. Child Development, 74, 205–220. Farmer, T. W., Leung, M. C., Pearl, R., Rodkin, P. C., Cadwallader, T. W., & Van Acker, R. (2002). Deviant or diverse peer groups? The peer affiliations of aggressive elementary students. Journal of Educational Psychology, 94, 611–620. Farmer, T. W., & Rodkin, P. C. (1996). Antisocial and prosocial correlates of classroom social positions: The social network centrality perspective. Social Development, 5, 174– 188. Farmer, T. W., Xie, H., Cairns, B. D., & Hutchins, B. C. (2007). Social synchrony, peer networks, and aggression in school. In P. H. Hawley, T. D. Little, & P. C. Rodkin (Eds.), Aggression and adaptation: The bright side to bad behavior (pp. 209–233). Mahwah, NJ: Lawrence Erlbaum Associates Publishers. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117– 140. Fiske, S. T. (1993). Controlling other people: The impact of power on stereotyping. American Psychologist, 48, 621–628. Gest, S. D., Farmer, T. W., Cairns, B. D., & Xie, H. (2003). Identifying children’s peer social networks in school classroom: Links between peer reports and observed interactions. Social Development, 12, 513–529. Gorman, A. H., Kim, J., & Schimmelbusch, A. (2002). The attributes adolescents associate with peer popularity and teacher preference. Journal of School Psychology, 40, 143–165. Harris, J. R. (1995). Where is the child’s environment? A group socialization theory of development. Psychological Review, 102, 458–489. Hartup, W. W. (1993). Adolescents and their friends. In B. Laursen (Ed.), Close friendships in adolescence (pp. 3–22). San Francisco, CA: Jossey-Bass. Hawley, P. H. (1999). The ontogenesis of social dominance: A strategy-based evolutionary perspective. Developmental Review, 19, 97–132. Hawley, P. H., Little, T. D., & Card, N. A. (2008). The myth of the alpha male: A new look at dominance-related beliefs and behaviors among adolescent males and females. International Journal of Behavioral Development, 32, 76–88. Jarvinen, D. W., & Nicholls, J. G. (1999). Adolescents’ social goals, beliefs about the causes of social success, and satisfaction in peer relations. Developmental Psychology, 32, 435– 441. Juvonen, J., & Murdock, T. B. (1995). Grade-level differences in the social value of effort: Implications for self-presentation tactics of early adolescents. Child Development, 66, 1694– 1705. Kruglanski, A. W., & Webster, D. M. (1991). Group members’ reactions to opinion deviates and conformists at varying degrees of proximity to decision deadline and of environmental noise. Journal of Personality and Social Psychology, 61, 212–225. LaFontana, K. M., & Cillessen, A. H. N. (2002). Children’s perceptions of popular and unpopular peers: A multimethod assessment. Developmental Psychology, 38, 635–647. LaFontana, K. M., & Cillessen, A. H. N. (2010). Developmental changes in the priority of perceived status in childhood and adolescence. Social Development, 19, 130–147. Lagerspetz, K. M. J., Björkqvist, K., & Peltonen, T. (1988). Is indirect aggression typical of females? Gender differences in aggressiveness in 11–12-year-old children. Aggressive Behavior, 14, 303–315. Latané, B. (1981). The psychology of social impact. American Psychologist, 36, 343–356. Lease, M. A., Musgrove, K. T., & Axelrod, J. L. (2002). Dimensions of social status in preadolescent peer groups: Likability, perceived popularity, and social dominance. Social Development, 11, 508–533. Lindenberg, S. (2001). Social rationality versus rational egoism. In J. Turner (Ed.), Handbook of sociological theory (pp. 635–668). New York: Kluwer/Plenum. Lindenberg, S. (2006). Prosocial behavior, solidarity, and framing processes. In D. Fetchenhauer, A. Flache, A. P. Buunk, & S. Lindenberg (Eds.), Solidarity and prosocial behavior: An integration of sociological and psychological perspectives (pp. 23–44). New York: Springer. Luke, D. A. (2004). Multilevel modeling. Thousand Oaks, CA: Sage. Maccoby, E. E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Harvard University Press. © Blackwell Publishing Ltd. 2011


Socialization of Physical and Social Aggression in Peer Groups 193 Maccoby, E. E. (2002). Gender and group process: A developmental perspective. Current Directions in Psychological Science, 11, 54–58. Maccoby, E. E., & Jacklin, C. N. (1974). The psychology of sex differences. Stanford, CA: Stanford University Press. Monahan, K., Steinberg, L., & Cauffman, E. (2009). Affiliation with antisocial peers, susceptibility to peer influence, and antisocial behavior during the transition to adulthood. Developmental Psychology, 45, 1520–1530. Ojanen, T., Grönroos, M., & Salmivalli, C. (2005). An interpersonal circumplex model of children’s social goals: Links with peer reported behavior and sociometric status. Developmental Psychology, 41, 699–710. Patterson, G. R., DeBaryshe, B. D., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329–335. Prinstein, M. J., & Cillessen, A. H. N. (2003). Forms and functions of adolescent peer aggression associated with high levels of peer status. Merrill-Palmer Quarterly, 49, 310– 342. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models (2nd ed.). Thousand Oaks, CA: Sage. Rodkin, P. C., Farmer, T. W., Pearl, R., & Van Acker, R. (2000). Heterogeneity of popular boys: Antisocial and prosocial configurations. Developmental Psychology, 36, 14–24. Rose, A. J., Swenson, L. P., & Waller, E. M. (2004). Overt and relational aggression and perceived popularity: Developmental differences in concurrent and prospective relations. Developmental Psychology, 40, 378–387. Rubin, K. H., Bukowski, W. M., & Parker, J. G. (2006). Peer interactions, relationships, and groups. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (pp. 571–645). Hoboken, NJ: John Wiley & Sons Inc. Rubin, K. H., & Coplan, R. J. (1992). Peer relationships in childhood. In M. H. Bornstein & M. E. Lamb (Eds.), Developmental psychology: An advanced textbook (pp. 519–578). Hillsdale, NJ: Erlbaum. Salmivalli, C., & Kaukiainen, A. (2004). ‘Female aggression’ revisited: Variable-and personcentered approaches to studying gender differences in different types of aggression. Aggressive Behavior, 30, 158–163. Steinberg, L. D., & Monahan, K. C. (2007). Age differences in resistance to peer influence. Developmental Psychology, 43, 1531–1543. Steinberg, L., & Silverberg, S. (1986). The vicissitudes of autonomy in early adolescence. Child Development, 57, 841–851. Tomada, G., & Schneider, B. H. (1997). Relational aggression, gender, and peer acceptance: Invariance across culture, stability over time, and concordance among informants. Developmental Psychology, 33, 601–609. Underwood, M. K., Scott, B. L., & Galperin, M. B. (2004). An observational study of social exclusion under varied conditions: Gender and developmental differences. Child Development, 75, 1538–1555. Vaughn, B. E., & Langlois, J. H. (1983). Physical attractiveness as a correlate of peer status and social competence in preschool children. Developmental Psychology, 19, 561– 567. Wentzel, K. R., & Asher, S. R. (1995). The academic lives of neglected, rejected, popular, and controversial children. Child Development, 66, 754–763. Xie, H., Cairns, R. B., & Cairns, B. D. (1999). Social networks and social configuration in inner-city schools: Aggression, popularity, and implications for students with EBD. Journal of Emotional and Behavioral Disorders, 7, 147–156. Xie, H., Cairns, R. B., & Cairns, B. D. (2001). Predicting teen motherhood and teen fatherhood: Individual characteristics and peer affiliations. Social Development, 10, 488–511. Xie, H., Cairns, R. B., & Cairns, B. D. (2002). The development of social aggression and physical aggression: A narrative analysis of interpersonal conflicts. Aggressive Behavior, 28, 341–355. Xie, H., Swift, D. J., Cairns, B. D., & Cairns, R. B. (2002). Aggressive behaviors in social interaction and development adaptations: A narrative analysis of interpersonal conflicts during early adolescence. Social Development, 11, 205–224. © Blackwell Publishing Ltd. 2011


194


Notes 1. In early adolescence, children often interact with peers in same-gender social groups (e.g., Cairns et al., 1998; Maccoby, 1998, 2002; Rubin et al., 2006). There are few mixed-gender groups in this period. The mechanism for cross-gender influence may be quite different from that of same-gender influence. Given these considerations, we focused on the socialization of aggression in the same-gender groups. Therefore, gender differences were tested using group gender rather than individual gender. 2. To test the validity of individual status measure, we examined its associations with leadership and perceived popularity. Participants were asked to nominate, from free recall, unlimited number of peers who best fit descriptors for items about leadership: ‘leader’ (‘this person gets chosen by the others as the leader. Other people like to have this person in charge’.), and about perceived popularity: ‘popular’ (‘some kids are very popular with their peers. That is, many classmates like to play with them or do things with them’.). For each student, the total number of peer nominations was calculated and standardized by school. Associations between individual status in a group and leadership and perceived popularity were highly significant (r = .21, p < .001 for being a leader; r = .22, p < .001 for being popular). This indicates that individual status is a robust measure of an individual’s prominence and power in his/her peer group.

© Blackwell Publishing Ltd. 2011
