Abstract:

Bullying is a group dynamic process, of which peer ecology is an essential context for the occurrence and development of bullying. Although existing studies have focused on peer factors that influence bullying, there is still limited research that systematically examines the development and changes in bullying-related behaviors from a group dynamics perspective. This study reviewed twenty social network studies and analyzed the development of bullying-related behaviors from structural and relational characteristics of social networks in the previous studies. As shown in Figure 1, these social networks include two parts: structures and relationships. Structural characteristics include group structure (e.g., classroom hierarchy) and individual position (e.g., embeddedness and betweenness). Relationship characteristics include static relationships (e.g., rejection and friendship) and dynamic relationships (e.g., selection and influence process).
Generally, these social network findings could be summarized as follows. First, as for the role of structural characteristics, it was found that classroom hierarchy (as an indicator of social prominence in the peer group) could directly predict the incidence and severity of bullying. Particularly, a classroom with a higher hierarchy tends to worsen the bullying situation, thus, all students in these classrooms are at a greater risk. At the same time, individual’s position in the network is a critical ecological marker for identifying those who are at risk for bullying, which could provide the possibility for early prediction and prevention. Second, as for the role of relationships in the network, on the one hand, these studies found that friendship increases the spread of bullying-related behaviors within the network, for example, bullying, victimization, and defense behaviors. On the other hand, these studies also demonstrated that bullying-related behaviors can format and maintain friendships in the networks, which bring together those with similar behaviors and exclude those with different behaviors. For example, bullies and defenders were unlikely to select victims as friends. Consequently, victimized individuals are excluded and isolated from various groups. By contrast, the mutual selection of bullies attracts them together and forms a "bully" gang. In this way, it polarizes bullying behaviors in groups and intensifies bullying-related behaviors from individual to group.
Future research could, first, explore the occurrence and development of bullying-related behaviors by measuring individuals' central location in the network (such as degree centrality, closeness centrality) and different relationship strengths, relationship scales, and relationship types in social networks from multiple perspectives. Second, future research should systematically and comprehensively examine the developmental changes in different bullying roles in networks to construct patterns of propagation and dissipation behaviors related to bullying roles and the patterns of mutual transformation among different roles. Meanwhile, the moderators and mediators should also be explored in future studies. Finally, future research should strengthen the collectivist culture and the importance of academic achievement, which could reveal the cultural and social environment imprint of bullying-related behaviors developing in China and provide a better empirical basis for bullying interventions from the perspective of group ecology.

Key words: bullying prevention, social network, group ecology, selection and influence effect, children and adolescents