<<

Popularity, Antipathy and Friendship Networks 1

Competition, , or Snobbism?

How Popularity and Friendships Shape Antipathy Networks of Adolescents

Christian Berger

Pontificia Universidad Católica de Chile

Jan Kornelis Dijkstra

University of Groningen

Please cite as:

Berger, C. & Dijkstra J.K. Competition, Envy, or Snobbism? How Popularity and

Friendships Shape Antipathy Networks of Adolescents, Journal of Research on Adolescence

(in press).

Popularity, Antipathy and Friendship Networks 2

Acknowledgements

This study was funded in part by a research fund granted to Christian Berger by the

Comisión Nacional de Desarrollo de Investigación Científica y Tecnológica, Chile

(FONDECYT project number 11110037).

Part of this research was funded by the Netherlands Organisation for Scientific

Research (NWO) Vernieuwingsimpuls (VENI) Project number 451-10-012 awarded to Jan

Kornelis Dijkstra (2010). Both authors made equal contributions to the manuscript. Popularity, Antipathy and Friendship Networks 3

Authors’ information

Christian Berger, Escuela de Psicología, Pontificia Universidad Católica de Chile.

Jan Kornelis Dijkstra, Department of Sociology and Interuniversity Center for Social

Science Theory and Methodology (ICS), University of Groningen, The Netherlands.

Correspondence regarding this manuscript should be send to Christian Berger.

Escuela de Psicología Pontificia Universidad Católica de Chile, Vicuña Mackenna 4860,

Macul, Santiago RM 782 0436, Chile, [email protected].

Popularity, Antipathy and Friendship Networks 4

Abstract

The aim of this study was to examine how status (popularity) and friendship relations

affected the development of adolescents’ dislike relations (i.e., antipathy networks) over time.

Three competing hypotheses were formulated about the role of status: antipathy relations

result from either similarity in status (‘competition-hypothesis’) or dissimilarity in status when lower status peers reject higher status peers (‘envy-hypothesis’) or vice versa

(‘snobbism-hypothesis’). Hypotheses were tested in a longitudinal sample of adolescents from Chile (5 th - 6th grade; 52% boys; N = 273) followed one year. Antipathy and friendship networks were examined simultaneously using longitudinal social network modelling

(SIENA). After structural network effects (e.g., reciprocity) and gender were controlled for, status dissimilarity was found to increase the likelihood of antipathy relationships,

particularly higher status adolescents were more likely to reject their lower status peers,

supporting the ‘snobbism-hypothesis’. Furthermore, best friends tended to agree upon which

peers to reject over time. No differences were found in the number of same-gender versus

cross-gender antipathies. Popularity, Antipathy and Friendship Networks 5

Competition, Envy, or Snobbism?

How Popularity and Friendships Shape Antipathy Networks of Adolescents

In adolescence, peer relations constitute an important social context for development.

Interpersonal relationships during this developmental phase take several forms, such as friendships, romantic relationships, bully-victim relationships, but also antipathetic relationships; relations of dislike. Antipathetic relationships appear to be a common form of social interaction in adolescents’ day-to-day lives (Güroglu, Haselager, Cornelis, Lieshout, &

Scholte, 2009), with approximately 35% of children having at least one (Card, 2010).

Nevertheless, researchers have just recently started to focus on this type of relationships. As argued by Card (2010), antipathies have been underinvestigated, even though they are highly relevant to understanding peer relations (Coie, Dodge, & Coppotelli, 1982; Parker & Gamm,

2003) and adolescents’ (mal)adjustment, such as externalizing problems, internalizing problems, and academic achievement (Card, 2010).

Traditionally, research has focused on peer rejection as an individual characteristic, measured by the number of dislike nominations adolescents receive from their peers

(Bierman, 2004; Newcomb, Bukowski, & Pattee, 1993). Bierman (2004) understands peer rejection as an actively disliking towards specific peers, but emphasizes that “it should always be kept in mind that peer rejection is an interactional process” (Bierman, 2004, p. 34).

Rejection should, therefore, be investigated from a relational perspective (Erath, Pettit,

Dodge, & Bates, 2009).

Recently researchers started to examine the relational nature of peer rejection by studying who dislikes whom on the dyadic level (Card, 2010). The term antipathy was coined for these dyadic dislike relations which can be either mutual or unilateral. Despite its relational nature, previous research has adopted an individual focus assessing correlates of being involved in antipathetic relationships (Parker & Gamm, 2003) or a dyadic approach Popularity, Antipathy and Friendship Networks 6 describing antipathies’ configurations (Berger, Rodkin, & Dijkstra, 2011). However, adolescents’ antipathies do not emerge in or solely in dyads but arise within the larger peer group and might consequently be affected by evaluations of other peers. For instance, within peer groups there might be more or less consensus about who is disliked, which in turn might individuals’ opinions about peers and steer conformity in antipathy relations. Similarly, other types of relations, such as friendships, might also influence who adolescents dislike (see also Huitsing et al., 2012).

This study aimed at contributing to an emerging field of research on antipathies by looking at these relations from a social network perspective. We were interested in the origins of antipathy relations and how they evolve over time by looking at the role of status

(popularity) and friendships in shaping these relations.

Theoretical explanations for links between popularity and antipathies

To understand what factors drive the emergence of antipathies, we used a goal- framing approach assuming that individuals are goal-oriented (Lindenberg, 2000, 2001,

2008). From a developmental perspective, gaining status and more specifically perceived popularity constitutes a central goal particularly in adolescence (Adler & Adler, 1998;

Ojanen, Grönroos, & Salmivalli, 2005; Pellegrini & Long, 2002), and is related to developmental processes associated with the definition of a social position within the peer culture (Potocnjak, Berger, & Tomicic, 2011; Rodkin, Farmer, Pearl, & Van Acker, 2000;

Xie, Li, Boucher, Hutchins, & Cairns, 2006). In this sense, interpersonal relationships help individuals to define their social position in the peer group by their choices of friends, but also with whom they have antipathy relations (Abecassis, 2003; Bierman, 2004; Cotterell,

2007); thus, interpersonal relationships might be functional to the achievement of their social goals, such as popularity (Salmivalli & Peets, 2009). Popularity, Antipathy and Friendship Networks 7

Research has already shown that perceived popularity is an important determinant for shaping friendships and the social structure in the peer group (Dijkstra, Berger, &

Lindenberg, 2011; Dijkstra, Cillessen, & Borch, 2012; Peters, Cillessen, Riksen-Walraven, &

Haselager, 2010). We extend to this by examining the role of perceived popularity (hereafter simply referred to as popularity) in the emergence of antipathies. Only two studies have examined directly the associations between perceived popularity and antipathetic relationships (Rodkin, Pearl, Farmer, & Van Acker, 2003; Witkow, Bellmore, Nishina,

Juvonen, & Graham, 2005). Witkow et al. (2005), featuring a cross-sectional design, found no association between antipathetic relationships and peer reported popularity, but did find a positive association between having a mutual antipathy and teacher reported popularity for girls. By contrast, Rodkin and colleagues (2003) found that boys and girls with multiple antipathies were less popular, whereas boys who lost same-sex antipathies over the school year increased in popularity. These studies, however, focused mainly on differences in popularity as a function of having antipathies or not, whereas we look at how popularity contributes to the development of antipathies over time. In view of status as an important goal in adolescence, and considering interpersonal relationships as functional to this mean, we expect that popularity plays a role in shaping antipathies over time. We developed three competing hypotheses.

As the achievement of status is a central goal in adolescence, but status is a positional asset which is always relative to peers (not everybody can hold a high-status position), competition for occupying high status positions within the peer group is likely to occur.

Anecdotal evidence comes from ethnographic studies describing how high status adolescents were aware of status increase of close peers that potentially threatened their position, which in turn evoked antipathetic (Adler & Adler, 1998; Merten, 1997). If true, antipathetic relationships may result from a pattern of competition, in which similarity in status fosters the Popularity, Antipathy and Friendship Networks 8 struggle for popular positions against peers, and thus underlies the development of these relationships. Accordingly, we formulated the following hypothesis: (1) competition- hypothesis : similarity in popularity leads to more competition for prevailing social positions, which in turn underlies the emergence of antipathies.

Alternatively, antipathetic relationships may also follow from dissimilarity in status.

Abecassis (2003) proposed that adolescents may choose whom to dislike based on personal characteristics and traits, choosing a peer who either possesses characteristics that one wishes to have or possesses characteristics that are disliked that one does not have. Applying this to dissimilarity in popularity, who dislikes whom may take two directions; lower status peers might dislike higher status peers or vice versa. Regarding the first option, adolescents who successfully gain status in the peer group might be envied by others who have failed to reach the goal of achieving high status. Still, high-status peers might also be attractive to lower status peers for affiliation as means of enhancing their own status (basking in reflected glory)

(Cialdini et al., 1976; Dijkstra, Cillessen, Lindenberg, & Veenstra, 2010), thus tempering the emergence of antipathy relations. Higher status peers in turn might want to keep lower status peers at a distance to avoid status loss and to keep unique social positions within the peer group (Dijkstra et al., 2012), resulting in arrogance and snobbism (Merten, 1997). This might cause antipathetic feelings towards lower status peers. Building on this, we formulated the following hypotheses: (2a) envy-hypothesis : dissimilarity in popularity leads to envy among lower status peers towards high popular peers, driving dislike and consequently favoring the development of antipathies; and (2b) snobbism-hypothesis : dissimilarity in popularity highlights the social dominance of higher popular peers, driving dislike of high popular peers towards lower popular peers.

Gender has also been proposed to play a significant role in who dislikes whom, particularly considering the gendered peer culture during early adolescence (Dijkstra, Popularity, Antipathy and Friendship Networks 9

Lindenberg, & Veenstra, 2007; Martin & Fabes, 2001). Previous research findings, however, are inconclusive as to whether boys or girls are more likely to be involved in antipathetic relationships, with some studies showing boys to have more antipathies (Güroglu et al., 2009;

Rodkin, Pearl, Farmer, & Van Acker, 2003), and others showing no differences (Abecassis,

Hartup, Haselager, Scholte, & van Lieshout, 2002; Berger et al., 2011; Parker & Gamm,

2003; Witkow et al., 2005). In lights of these findings, we took the role of gender into account by examining if gender affected the likelihood of giving and receiving dislike nominations, and whether same-gender or cross-gender antipathies were more prevalent.

To examine the roles of popularity and gender we examined antipathy relations as a larger social network. Although antipathies appear at first glance to be a dyadic evaluation of disliked peers, these relations might be steered by the extent to which adolescents are involved in antipathies by either giving or receiving nominations for dislike, and subsequently attracting more antipathies from other peers over time. As argued by Bierman

(2004), reputational biases are developed within peer groups regarding rejected peers, and affect the ways in which adolescent perceive, evaluate, feel and act regarding them. Hence, structural characteristics of the peer network such as its density, the likelihood for nominations to be reciprocated, and the self-reinforcing effect of either giving or receiving nominations as disliked may play a role in shaping antipathetic relationships; therefore, we controlled for these network effects. In doing so, we gain more reliable estimates of the impact of popularity on the development of antipathies.

In addition, while considering antipathies as a larger social network we also considered the interplay with friendship networks, utilizing a multiple cross-network approach. Peer networks are not encapsulated but are affected by other networks that co- occur. From this perspective, friendship networks are likely to shape—and be shaped by— dislike networks within a certain peer context (Laursen et al., 2010). We examined to what Popularity, Antipathy and Friendship Networks 10 extent friendship networks affected the emergence of antipathies and vice versa by looking at the interplay between both networks. Several studies on friends’ homophily showed that adolescents tend to conform with their friends regarding their attitudes and behaviors (Berger

& Rodkin, 2012; Fergusson, Vitaro, Wanner, & Brendgen, 2007; Jaccard, Blanton, & Dodge,

2005; Mercken, Snijders, Steglich, Vartiainen, & Vries, 2010; Valenzuela & Ayala, 2011).

Building on balance theory (Heider, 1958) in which it is argued that individuals strive for psychological balance in relationships, we expected that the antipathies of friends also become your antipathies. Hence, we expect the following multiple network effects: (3a) friends’ agreement hypothesis : friends would be likely to agree about their antipathies. The reverse might also be the case, and constitutes our final hypothesis: (3b) from antipathies to agreement hypothesis : sharing similar antipathies facilitates the formation of friendship relations.

The present study aimed at broadening the understanding of how antipathies develop over time, constituting to our knowledge the first study in which antipathies have been examined in a developmental social network framework. Specifically, this study focused on how popularity shapes antipathies, while taking into account the role of structural network effects, gender, and the interplay of antipathies with friendship networks. In these friendship networks we also examined the effects of gender and popularity, resulting in a full multiple network model.

Method

Participants

Participants were part of a larger study on peer relations. In total, 273 fifth and sixth graders (134 fifth and 139 sixth graders; 52.0% boys, age range 10 -12) from two urban schools in metropolitan Santiago, Chile, were included in the study (see also Dijkstra et al.,

2011). These schools were private but received a public subsidy, representing the majority of Popularity, Antipathy and Friendship Networks 11 the Chilean school population. Both schools were average in terms of their family income, and were located in low to middle socioeconomic status neighborhoods. In each school, all students were approached to participate in the data collection. Active consent was gathered from all students and their parents. For 25 students no information was available at Wave 1, and were coded as missing. Attrition analyses showed that participants who were present at both assessments were more likely to be disliked at Wave 2 than participants who were present only at the second assessment. Participants who participated in both waves had more nominations as best friend as compared to their peers who participated in only one wave. No differences were found regarding popularity. Ethnic information was not collected due to the homogeneity in the ethnic configuration of the Chilean society. In the Chilean educational system classmates remain together for at least their whole primary education (grades 1 st to

8th ). Classrooms are fixed, that is, most of the times students do not change classroom.

Procedure

Participants were surveyed from June to August (middle of the academic year), and reassessed during the same months one year later. Surveys were completed during regular class hours through group administration, taking 45 minutes per classroom. Participants were assured that their answers would be kept confidential, and were told that they were not allowed to talk and that they could stop participating at any time. Measures, consent protocols and procedures to protect the confidentiality and rights of all participants were approved by the Institutional Review Board of the local university and by the principals of both schools involved in this research.

Measures

Antipathy networks . Participants were allowed to nominate up to six classmates they disliked. On average, participants nominated 2.28 classmates as disliked at Wave 1 and 2.31 at Wave 2. These dislike nominations were used to determine antipathy networks using Popularity, Antipathy and Friendship Networks 12 adjacency matrices, containing information on whether a dislike relation was absent (zero) or present (one). An overall adjacency matrix was used, including all eight classroom networks.

Structural zeros between classroom networks were used to indicate that participants were not able to nominate peers from other classrooms.

Friendship networks . Participants were also asked to write down the names of six children who they considered as best friends in their classroom. On average, participants selected 2.53 friends at Wave 1 and 2.90 at Wave 2. Friendship nominations were used to assess the friendship networks. Again, an overall network was computed with structural zeros between classroom networks.

Popularity . Participants were allowed to nominate up to six classmates they considered to be ‘popular’ and ‘not popular’. Following LaFontana and Cillessen’s procedure

(LaFontana & Cillessen, 2002), popularity was calculated by subtracting peer nominations as

“not popular” from peer nominations as “popular”. Scores were then standardized within class and z-standardized across networks. As recommended (Snijders, Steglich, & Van de

Bunt, 2010), status was subsequently transformed into a categorical variable, that is, a 4-point scale, using increments of .50 of the continuous z-score as cut-off points. This categorization nicely captured the distribution of popularity in our sample and included sufficient variability to estimate changes in popularity, also yielding the best model fit.

Analyses

We examined the development of antipathy networks using longitudinal social network modeling (R SIENA 4.0). In addition to the antipathy networks, we also included the best friendship networks to examine the interplay between these multiple networks. For both networks, we examined the following structural network effects: density , the number of outgoing ties, and, therefore, the density of the network; reciprocity , the extent to which nomination choices are reciprocated. For the best friendship network we also examined Popularity, Antipathy and Friendship Networks 13 transitive triplets ; the tendency of individuals to be friends with the friends of their friends.

For the antipathy networks we included two additional network effects: indegree-dislike and outdegree-dislike. Indegree-dislike reflects the tendency for respondents with a high number of incoming ties (nominations received) to attract extra incoming dislike nominations at

Wave 2. This effect can be interpreted as a reinforcement effect of rejection. Outdegree- dislike indicates the extent to which respondents with a high number of outgoing ties

(nominations given) at Wave 1 tend to receive more dislike nominations over time (i.e., effect of dislike nominations given on receiving dislike nominations). These two effects are usually referred to as indegree-popularity and outdegree-popularity in SIENA. However, to avoid confusion with our variable of main – popularity – we labeled these effects as indegree-dislike and outdegree-dislike.

For both the antipathy and friendship networks, we modeled the effects of gender and popularity on nominations given (ego-effects) and received (alter effects) , and the extent to which nominations were likely to be same-gender ( same-gender selection ) and for same- status peers ( popularity-similarity selection ). Finally, we included two multiple network effects. The effect of friends to agreement on antipathies reflects the likelihood of friends at

Wave 1 agreeing in their outgoing ties for dislike at Wave 2. The effect from antipathy

agreement on friendships models whether sharing same antipathies underlies friendship

formation. Both effects reflect the interplay between multiple networks by showing how the

antipathy and friendship networks shape each other over time. We also included the basic

effect of antipathies on friendships and vice versa in the model. Both effects were fixed at a

high negative value of minus 5, because these effects were difficult to estimate (due to that it

is very unlikely that a dislike nomination would result in a friendship nomination).

To further examine the selection effects in the antipathy networks at different levels of

popularity, we additionally constructed an ego-alter selection table (Snijders, Steglich, Popularity, Antipathy and Friendship Networks 14

Schweinberger, & Huisman, 2007). This table indicates for each level of popularity the propensity of adolescents (in the rows) to select peers in any of the other four categories (in the columns) as disliked. For the ego-alter table we used the unstandardized estimates from the SIENA model for popularity ego, popularity alter, and popularity similarity effect, following standard SIENA recommendations (Ripley, Snijders, & Preciado, 2011; Snijders et al., 2007).

Results

Descriptive statistics and correlations

Correlations between outdegree and indegree measures of our main variables at both

assessments (Table 1) showed that adolescents who gave more dislike nominations were

more likely to be nominated as disliked; seemingly, participants who were nominated as

friends were more likely to reciprocate friendship nominations and to be nominated as

popular. All outdegree measures were positively associated with each other. Only for the first

assessment participants nominated as friends were less nominated as disliked, and

participants who were nominated as popular were more likely to give dislike nominations. In

the second assessment participants who received nominations as friends gave more dislike

nominations; seemingly, participants who were nominated as popular gave more friendship

nominations. Girls were more likely to nominate others as disliked and less likely to receive

nominations as popular. On the second assessment girls also gave more popular nominations.

Friendship was unrelated to gender.

Regarding descriptive network statistics, the average number of nominations given for

dislike was somewhat lower than for the best friendship network, which was also reflected by

less dense antipathy networks than friendship networks (Table 2). Furthermore, antipathy

relations were more often asymmetric than friendship relations. Between both time points

more antipathies were created than resolved with a relatively small proportion of stable Popularity, Antipathy and Friendship Networks 15 antipathies. Although friendship relations tended to be more stable over time than antipathies, the relatively low Jaccard-indices suggest that friendship relations were also subject to change. Considering the increase of friendship relations between both time points, we followed the suggestion of Snijders et al. (2010) by also calculating the proportion of relationships that remained relative to the number of stable and dissolved relationships. These proportion were acceptable (> .30) (see Snijders et al., 2010).

Testing hypotheses

The structural network effects for the best friendship network (Table 3) revealed that respondents were likely to nominate fewer than half of their peers as friends as indicated by the negative density effect. Furthermore, friendship nominations were likely to be reciprocated over time (positive reciprocity effect) and friends of friends tended to become friends as well (positive transitive triplets effect).

There was a strong preference for same-gender friends as well as for friends who were

similar in popularity. Additionally, the positive popularity alter effect indicates that

popularity increased the number of best friend nominations received.

With regard to the structural network effects for the antipathy network, it appeared

that antipathy networks were relatively sparse, as indicated by the negative density parameter.

The reciprocity effect revealed that dislike nominations tend to be reciprocated over time. We

also found that higher numbers of nominations received at the first wave increased the

likelihood of receiving extra nominations one year later, as indicated by the indegree-dislike

effect. No effect was found for respondents with a high number of dislike nominations given

at Wave 1 to receive more dislike nominations over time (outdegree-dislike ).

One significant gender effect was found: male respondents gave fewer dislike

nominations than girls (gender ego ). No significant effect was found for same-gender Popularity, Antipathy and Friendship Networks 16 selection , suggesting that dislike nominations were equally likely to be same-gender as cross- gender.

With regard to popularity, we found a marginally significant effect for popularity alter, suggesting that higher popular respondents were less likely to be nominated by peers as disliked (popularity alter ). We also found a negative popularity-selection effect, meaning that

dissimilarity in popularity resulted in antipathy relationships over time.

To untangle the direction of who nominates whom as dislike, that is, whether low-

status adolescents were more likely to dislike higher popular peers or vice versa, we

calculated the selection effects for each category of popularity (see Table 4). This table

indicates the strength of dislike from the interplay between popularity of the adolescent and

popularity of the peer. As can be seen, the likelihood of selecting same-popular peers as

disliked decreased when popularity of the respondents increased (see diagonal in Table 4). It

also appeared that higher popular adolescents were more likely to dislike lower popular peers

than the other way around. This is revealed by a stronger effect of antipathy feelings

regarding low popular peers when being popular (.26) than antipathy feelings regarding high

popular peers when being low in popularity (.08). This supports the snobbism-hypothesis that

dissimilarity in popularity steers antipathies mainly from higher popular peers towards lower

popular peers.

Looking at the multiple network effects, findings were only in line with the friend- agreement hypothesis; friends tended to agree on their dislike nominations over time, targeting the same peers as disliked. This effect indicates that antipathies partially arise from agreement among friends regarding whom they dislike. No effect was found for antipathies resulting in friendship relations over time.

Discussion Popularity, Antipathy and Friendship Networks 17

We examined how status affected changes in antipathies from a social network perspective by taking into account the role of the friendship network at the same time. In doing so, we also took into account the roles of gender and the network structure of antipathies by investigating reciprocity as well as whether involvement in antipathy relations, either as nominator or as nominee, gave rise to more antipathies over time (indegree-dislike and outdegree dislike effects). Theoretically, we formulated three competing hypotheses. On the one hand, we argued that adolescents who are similar in status might perceive each other as competitors, leading to antipathy relations (‘competition-hypothesis’). This is in line with

Erath and colleagues (2009), who argued that adolescents might respond to competitive interactions like a symmetric escalation to generate dissimilarity. On the other hand, we argued that dissimilarity in status might trigger antipathies either through low-status peers rejecting higher status peers (‘envy-hypothesis’) or higher status peers rejecting lower status peers (‘snobbism- hypothesis’). Previous studies have found partial support for the dissimilarity hypothesis (Nangle, Erdley, Zeff, Stanchfield, & Gold, 2004; Nangle, Erdley, &

Gold, 1996), showing that children were more likely to dislike peers who were different instead of similar in behavioral profiles and likeability.

Our findings showed that mainly status dissimilarity increased the likelihood of antipathy relations, particularly from high popular peers towards lower popular peers, lending support to the snobbism-hypothesis. Adolescents, who are lower in popularity might still try to seek affiliation with more popular peers. High status peers, however, can be more selective in view of the higher number of nominations they receive as best friend. Similar patterns emerged in a longitudinal study on two middle schools in the United States, revealing a strong tendency of high status adolescents to affiliate with each other and avoid lower status peers as friends (Dijkstra et al., 2012). Popularity, Antipathy and Friendship Networks 18

Although high status adolescents might have feelings of competition among each other (Merten, 1997), this did not result in antipathetic relationships in our study. One possible explanation is that antipathies between peers with similar popularity level imply higher risks for the individual. Peers with similar social status (and power) levels could be more threatening than lower status peers to one’s social position, which makes disliking a dangerous strategy. Hence, popular adolescents might therefore prefer to encapsulate their competitors by being friends with each other, and extend competition towards the outgroup; lower status peers. An interesting avenue for future research is to examine such ingroup- outgroup processes based on status and the extent to which they give rise to both positive and negative types of peer relations within and between groups.

Still, it might also be that having a similar status avoids showing active dislike to each other, but leads to more subtle ways of peer antipathetic relations. For instance, previous studies have shown that negative judgments towards peers are motivated by self- enhancement goals, particularly when contrast implies a threat to a person’s self-worth

(Beauregard & Dunning, 1997). Further research could address these issues assessing different forms of negative affect among peers, such as avoidance, victimization, denial, and gossiping.

Previous studies showed no relation between having antipathies and social status in terms of popularity and social visibility (Rodkin et al., 2003; Witkow et al., 2005). Our findings, however, revealed that status shapes the antipathy network over time. One explanation is that we examined the impact of status on the development of antipathies within a social network perspective. Although antipathies could be approached as an individual trait

(rejection) or at the dyadic level, our social network approach adds to these perspectives by showing that antipathies are indeed embedded in a larger network. This was supported by the finding that adolescents who received a large number of dislike nominations at wave one Popularity, Antipathy and Friendship Networks 19 attracted additional dislike nominations one year later from other peers in the network. These findings underlie the fact that involvement in antipathies seems to be reinforced within the peer group over time.

Along with the social network perspective, a novel feature of this study was the inclusion of multiple networks within the adolescent peer ecology. Antipathetic relationships do not unfold spontaneously; by the contrary, they are influenced by other relational dynamics. We were particularly interested in the interplay between antipathies and friendships. Whereas the friends-agreement hypothesis was supported by showing that friends tend to have the same antipathies over time, the reversed pattern that sharing antipathy relations creates friendship relations was not found. Apparently, negative relations do emerge from positive ones, but not the other way round. This makes sense considering that in an already existing friendship conformity in opinions might arise due to influence processes. However, in the absence of a positive relationship only sharing the same antipathy might be less powerful to establish a relationship. Yet, this could still be worthwhile to explore in other types of relationships, particularly bully-victim relationships (Huitsing, et al.,

2012). Being target of the same bully could give rise to friendships among victims. The integration of co-occurring, multiple networks within the same analyses opens a new perspective on studying peer relations.

There are some limitations we should acknowledge. First, we only focused on the role of status and gender in this study. Future research should further explore other characteristics and behaviors that might also affect the emergence of antipathies, either directly or through links with popularity or other forms of social status within the peer group. A growing body of studies has shown the importance of popularity for the structure and development of friendship relations (Dijkstra et al., 2011; 2012). This study adds to these studies by showing the importance of popularity in the emergence of antipathies. More research on antipathies Popularity, Antipathy and Friendship Networks 20 should not only focus on potential risks for engagement and development of antipathies, but also factors that protect adolescents against antipathetic relationships. For instance, prosocial behavior has consistently been shown to reduce peer rejection (Dijkstra et al., 2007;

Haselager, Cillessen, Van Lieshout, Riksen-Walraven, & Hartup, 2002; Newcomb,

Bukowski, & Pattee, 1993) and antipathetic relationships (Card, 2010).

Second, this study was conducted in a relatively small sample across two time points with a one-year time interval. Between both time points, the stability of friendship relations was relatively low. Ideally, networks should show more stability to meet the assumption of

SIENA that change processes are gradual. Moreover, our networks were limited to the classroom, excluding potential important peer relations within grade and school, and forcing the inclusion of structural zeros between classrooms. Nevertheless, within these relatively small networks our study already revealed interesting patterns concerning the role of status in shaping antipathies. Future research might profit from larger samples with more time points in order to gain a better understanding of the developmental processes over time.

Third, we only focused on the network dynamics of antipathies. It would be interesting for future research also to look at behavioral dynamics and examine to what extent antipathy relations influence the development of behaviors over time. Until now, research on adolescents’ social networks has almost solely focused on positive relations, most prominently, friendship relations, showing that friends influence often results in increased similarity in a wide variety of behaviors over time (Veenstra & Dijkstra, 2011). Antipathies might have the opposite effect with adolescents who want to distinguish themselves from the peers they dislike by becoming more dissimilar to them.

Fourth, the study featured a sample from Chile. Yet, most research on adolescents’ peer relations has been conducted in United States and Europe. Although Chile is considered as one of the most westernized countries in Latin America, the generalizability of these Popularity, Antipathy and Friendship Networks 21 findings might therefore be taken with some caution. Nevertheless, using data from non-

Western countries is important for understanding how peer relations function and impact adolescent development across different countries and cultures (Chaux, 2011).

In this regard, our study is an important first step in disentangling the dynamics that

give rise to antipathies among adolescents by looking at the role of popularity and at the co-

occurrence of both positive and negative peer networks. Popularity, Antipathy and Friendship Networks 22

Table 1

Correlations between Main Variables (N = 273) at the first (below diagonal) and second assessment (above diagonal)

1 2 3 4 5 6 7

1. Indegree dislike - .25** -.06 .04 .04 .11 -.06

2. Outdegree dislike .23** - .21** .39** .04 .48** .16*

3. Indegree friendship -.22** .12 - .43** .45** .26** -.02

4. Outdegree friendship .06 .40** .26** - .19** .45** -.03

5. Indegree popularity -.04 .16* .46** .11 - .17** -.16*

6. Outdegree popularity .07 .47** .18** .54** .13* - .13*

7. Gender (1 = male) -.12 .19** -.02 .01 -.14* .08 -

Note . For all measures we used proportion scores.

* p < .05; ** p < .01

Popularity, Antipathy and Friendship Networks 23

Table 2

Descriptive Statistics for Friendship and Antipathy Networks (N = 273)

Friendship Networks Antipathy Networks

Wave 1 Wave 2 Wave 1 Wave 2

Density 1 .009 .011 .008 .009

Average degree 2.53 2.90 2.28 2.31

Number of ties 678 784 609 615

Mutual 144 200 56 55

Asymmetric 314 335 428 482

Missing fraction .02 .01 .02 .01

Tie changes

Absence of tie (0 0) 71162 70359

Creating tie (0 1) 457 437

Resolving tie (1 0) 428 362

Stable tie (1 1) 208 121

Jaccard-index 2 .19 .13

Note . 1Density reflects the proportion of friendship relations relative to the total number of

possible relations.

2Jaccard-index indicates the proportion of stable relations of the total number of created, resolved and stable relations. Popularity, Antipathy and Friendship Networks 24

Table 3

Results SIENA analyses (N = 273)

Est. SE

Friendship Network

Structural Network Effects

Density -2.23* .10

Reciprocity 1.23* .16

Transitive triplets .23* .05

Gender Effects

Male alter -.09 .09

Male ego .14 .12

Same-gender selection .75* .07

Popularity Effects

Popularity alter .13* .04

Popularity ego -.05 .04

Popularity similarity selection .38* .10

Antipathy Network

Structural Network Effects

Density -1.88* .12

Reciprocity .52* .20

Indegree-dislike .07* .02

Outdegree-dislike .02 .05

Gender Effects

Male alter .06 .08

Male ego -.22* .08 Popularity, Antipathy and Friendship Networks 25

Same-gender selection .02 .06

Popularity Effects

Popularity alter -.06+ .03

Popularity ego -.01 .03

Popularity similarity selection -.32* .12

Multiple-Network Effects

Friends to agreement (friends agree upon .45* .09

shared antipathies)

From antipathy agreement (sharing same .07 .18

antipathies creates friendship)

+ p <.10 / * p < . 05 Popularity, Antipathy and Friendship Networks 26

Table 4

Selection Effects for Antipathy Networks of Adolescent Popularity on Peer Popularity (N =

273)

Popularity of Peer

1 2 3 4

1 -0.05 -0.01 0.04 0.08

Popularity of 2 0.05 -0.12 -0.08 -0.03

Adolescent 3 0.15 -0.02 -0.19 -0.14

4 0.26 0.09 -0.09 -0.26

Popularity, Antipathy and Friendship Networks 27

References

Abecassis, M. (2003). I hate you just the way you are: exploring the formation, maintenance, and need for enemies. New Directions for Child and Adolescent Development, 102 , 5- 22. doi:10.1002/cd.86

Abecassis, M., Hartup, W. W., Haselager, G. J. T., Scholte, R. H. J., & Van Lieshout, C. F. M. (2002). Mutual antipathies and their significance in middle childhood and adolescence. Child Development , 73 (5), 1543-1556. doi: 10.1111/1467-8624.00489

Adler, P. A., & Adler, P. (1998). Peer power: Preadolescent culture and identity . New Brunswick: Rutgers University Press.

Beauregard, K., & Dunning, D. (1997). Turning up the contrast: Self-enhancement motives prompt egocentric contrast effects in social judgments. Journal of Personality and Social , 74 (3), 606-621. doi: 10.1037/0022-3514.74.3.606

Berger, C., & Rodkin, P. C. (2012). Group influences on individual aggression and prosociality: Early adolescents who change peer affiliations. Social Development , 21 (2), 396-413. doi:10.1111/j.1467-9507.2011.00628.x

Berger, C., Rodkin, P. C., & Dijkstra, J. K. (2011). Antipathetic relationships among adolescents : Exploring prevalence , gender differences , and stability in the United States and Chile. Anales de Psicologia , 27 , 783-790.

Bierman, K. L. (2004). Peer rejection: Developmental processes and intervention strategies . New York: Guilford Press.

Card, N. (2010). Antipathetic relationships in child and adolescent development: a meta- analytic review and recommendations for an emerging area of study. Developmental psychology , 46 (2), 516-529. doi:10.1037/a0017199

Chaux, E. (2011). Multiple Perspectives on a Complex Problem : Comments on Five Studies on School Violence. Psykhe, 20 (2), 79-86. doi: 10.4067/S0718- 22282011000200007

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 (3), 366-375. doi: 10.1037/0022-3514.34.3.366

Coie, J., Dodge, K., & Coppotelli, H. (1982). Dimensions and types of social status: A cross- age perspective. Developmental Psychology , 18 (4), 557-570. doi: 10.1037/0012- 1649.18.4.557

Cotterell, J. (2007). Social networks in youth and adolescence (2nd ed.). New York: Routledge. Popularity, Antipathy and Friendship Networks 28

Dijkstra, J. K., Berger, C., & Lindenberg, S. (2011). Do physical and relational aggression explain adolescents’ friendship selection? The competing roles of network characteristics, gender, and social status. Aggressive Behavior , 37 , 417-429. doi:10.1002/ab.20402

Dijkstra, J. K., Cillessen, A. H. N., Lindenberg, S., & Veenstra, R. (2010). Same-Gender and Cross-Gender Likeability: Associations With Popularity and Status Enhancement: The TRAILS study. The Journal of Early Adolescence , 30 (6), 773-802. doi:10.1177/0272431609350926

Dijkstra, J.K., Lindenberg, S., & Veenstra, R. (2007). Same-gender and cross-gender peer and peer rejection and their relation to bullying and helping among preadolescents: comparing predictions from gender-homophily and goal-framing approaches. Developmental Psychology , 43 (6), 1377-1389. doi:10.1037/0012- 1649.43.6.1377

Dijkstra, J. K., Cillessen, A. H. N., & Borch, C. (2012). Popularity and adolescent friendships networks: Selection and influence dynamics. Developmental Psychology , in press. doi: 10.1037/a0030098

Erath, S. A., Pettit, G. S., Dodge, K. A., & Bates, J. E. (2009). Who Dislikes Whom, and For Whom Does It Matter: Predicting Aggression in Middle Childhood. Social Development , 18 (3), 577-596. doi:10.1111/j.1467-9507.2008.00497.x

Fergusson, D. M., Vitaro, F., Wanner, B., & Brendgen, M. (2007). Protective and compensatory factors mitigating the influence of deviant friends on delinquent behaviours during early adolescence. Journal of Adolescence , 30 (1), 33-50. doi:10.1016/j.adolescence.2005.05.007

Güroglu, B., Haselager, G. J. T., Cornelis, F. M., Lieshout, V., & Scholte, R. H. J. (2009). Antagonists in mutual antipathies: A person-oriented approach. Adolescence , 19 (1), 35- 46. doi: 10.1111/j.1532/7795.2009.00580.x

Haselager, G. J. T., Cillessen, A. H. N., Van Lieshout, C. F. M., Riksen-Walraven, J. M. a., & Hartup, W. W. (2002). Heterogeneity among peer-rejected boys across middle childhood: Developmental pathways of social behavior. Developmental Psychology , 38 (3), 446-456. doi:10.1037//0012-1649.38.3.446

Heider, F. (1958). The psychology of interpersonal relations . New York: Wiley.

Jaccard, J., Blanton, H., & Dodge, T. (2005). Peer influences on risk behavior: an analysis of the effects of a close friend. Developmental Psychology , 41 (1), 135-47. doi:10.1037/0012-1649.41.1.135

LaFontana, K. M., & Cillessen, A. H. N. (2002). Children’s perceptions of popular and unpopular peers: A multimethod assessment. Developmental Psychology , 38 (5), 635- 647. doi:10.1037//0012-1649.38.5.635 Popularity, Antipathy and Friendship Networks 29

Laursen, B., Bukowski, W. M., Nurmi, J.-eri, Marion, D., Salmela-aro, K., & Kiuru, N. (2010). Opposites detract: Middle school peer group antipathies. Journal of Experimental Child Psychology , 106 (4), 240-256. doi:10.1016/j.jecp.2010.03.001

Lindenberg, S. (2000). ‘James Coleman’. In G. Ritzer (Ed.), The Blackwell Companion to Major Social Theorists (pp. 513-544). Oxford: Blackwell Publishers.

Lindenberg, S. (2001). Social rationality versus rational egoism. In J. Turner (Ed.), Handbook of Sociological Theory (pp. 635-668). New York: Kluwer Academic/Plenum.

Lindenberg, S. (2008). Social Rationality, Semi-Modularity and Goal-Framing: What Is It All About? Analyse & Kritik , 30 , 669-687.

Martin, C. L., & Fabes, R. A. (2001). The stability and consequences of young children’s same-sex peer interactions. Developmental Psychology , 37 (3), 431-446. doi:10.1037//0012-1649.37.3.431

Mercken, L., Snijders, T. A. B., Steglich, C., Vartiainen, E., & Vries, H. D. (2010). Dynamics of adolescent friendship networks and smoking behavior. Social Networks , 32 , 72-81. doi:10.1016/j.socnet.2009.02.005

Merten, D. E. (1997). The meaning of meanness: Popularity, competition, and conflict among junior high school girls. Sociology of Education , 70 (3), 175-191.

Nangle, D., Erdley, C. A., Zeff, K. R., Stanchfield, L. L., & Gold, J. A. (2004). Opposites do not attract : Social status and behavioral-style concordances and discordances among children and the peers who like or dislike them. Journal of Abnormal Child Psychology , 32 (4), 425-434. doi: 10.1023/B:JACP.0000030295.43586.32

Nangle, D., Erdley, C., & Gold, J. (1996). A reflection on the popularity construct: The importance of who likes or dislikes a child. Behavior Therapy , 27 (3), 337-352. doi: 10.1016/S0005-7894(96)80021-9

Newcomb, A., Bukowski, W. M., & Pattee, L. (1993). Children’s peer relations: A meta- analytic review of popular, rejected, neglected, controversial, and average sociometric status. Psychological Bulletin , 113 , 99-128. doi: 10.1037/0033-2909.113.1.99

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 (5), 699-710. doi:10.1037/0012-1649.41.5.699

Parker, J. G., & Gamm, B. K. (2003). Describing the dark side of preadolescents’ peer experiences: Four questions (and data) on preadolescents' enemies. New Directions for Child and Adolescent Development, 102 , 55-72. doi: 10.1002/cd.89

Pellegrini, A. D., & Long, J. D. (2002). A longitudinal study of bullying, dominance, and victimization during the transition from primary school through secondary school. British Journal of Developmental Psychology , 20 (2), 259-280. doi:10.1348/026151002166442 Popularity, Antipathy and Friendship Networks 30

Peters, E., Cillessen, A. H. N., Riksen-Walraven, J. M., & Haselager, G. J. T. (2010). Best friends’ preference and popularity: Associations with aggression and prosocial behavior. International Journal of Behavioral Development , 34 (5), 398-405. doi:10.1177/0165025409343709

Potocnjak, M., Berger, C., & Tomicic, T. (2011). A relational approach to school violence among peers in Chilean adolescents: Adolescent perspective about intervening factors. Psykhe , 20 (2), 39-52. doi: 10.4067/S0718-22282011000200004

Ripley, R., Snijders, T. A. B., & Preciado, P. (2011). Manual for SIENA version 4.0. Oxford: University of Oxford, Department of Statistics; Nuffield College.

Rodkin, P. C., Farmer, T. W., Pearl, R., & Van Acker, R. (2000). Heterogeneity of popular boys: antisocial and prosocial configurations. Developmental Psychology , 36 (1), 14-24. doi: 10.1037/0012-1649.36.1.14

Rodkin, P. C., Pearl, R., Farmer, T. W., & Van Acker, R. (2003). Enemies in the gendered societies of middle childhood: prevalence, stability, associations with social status, and aggression. New directions for Child and Adolescent Development, 102 , 73-88. doi:10.1002/cd.90

Salmivalli, C., & Peets, K. (2009). Pre-adolescents’ peer-relational schemas and social goals across relational contexts. Social Development , 18 (4), 817-832. doi:10.1111/j.1467- 9507.2008.00515.x

Snijders, T., Steglich, C., Schweinberger, M., & Huisman, M. (2007). Manual for SIENA version 3 . Groningen: Groningen: University of Groningen, ICS. Oxford: University of Oxford, Department of Statistics.

Snijders, T., Steglich, C., & Van de Bunt, G. (2010). Introduction to stochastic actor-based models for network dynamics. Social Networks , 32 , 44-60. doi: 10.1016/j.socnet2009.02.004

Valenzuela, E., & Ayala, C. (2011). Homophily, selection, and influence in a longitudinal study of drugs in school population. Psykhe , 20 , 101-114. doi: 10.4067/S0718- 22282011000200009

Veenstra, R., & Dijkstra, J. K. (2011). Transformations in adolescent peer networks. In B. Laursen & W. A. Collins (Eds.), Relationships pathways: From adolescence to young adulthood (pp. 135-154). Los Angeles: Sage.

Witkow, M., Bellmore, A., Nishina, A., Juvonen, J., & Graham, S. (2005). Mutual antipathies during early adolescence: More than just rejection. International Journal of Behavioral Development , 29 (3), 209-218. doi:10.1080/01650250444000513

Xie, H., Li, Y., Boucher, S. M., Hutchins, B. C., & Cairns, B. D. (2006). What makes a girl (or a boy) popular (or unpopular)? African American children’s perceptions and developmental differences. Developmental Psychology , 42 (4), 599-612. doi:10.1037/0012-1649.42.4.599