Network features that moderate the association between types of social aggression and status in children's peer groups

Michele Lease, Dawn Robinson, Emily Lee, Jamilia Blake


The focus of childhood bullying research has shifted from direct (physical, verbal) to more subtle forms of indirect aggression that inflict harm through damaging relationships, through social exclusion, including use of gestures, or by damaging reputations (Murray-Close et al., 2016; Salmivalli & Peets, 2018). One understudied form of social aggression is the use of nonverbal gestures to convey disdain, making someone feel excluded or unwelcome (Underwood, 2004). Although rarely included on assessments (Voulgaridou & Kokkinos, 2019), some research has shown it to be distinct from other types of social aggression (Blake et al., 2011). Because it is potentially less (socially) costly, it is also likely to become a more common form of social aggression across time. The objective of the current research is to identify network features (e.g., proximity to the aggressor versus having a tie with the aggressor that can be manipulated) that help distinguish nonverbal aggression from other types of social aggression. We are ultimately interested in the conditions that contribute to the effective use of various forms of social aggression to influence peers and to assert social control within the peer “hanging-out” network. In particular, we examine whether network position moderates the effect of social aggression on social status. Participants in 21 4th-5th grade classrooms nominated classmates as nonverbal (“acts bored and sighs when someone they don't like is talking”; “stare at a kid and acts disgusted...when don't like the kid”), social (“tries to keep certain people from being in their group”), and /or relationally (“tell the child will stop liking them unless do what they say”) aggressive as well as having social control (“decides who gets to be in the popular/in crowd”) and influence (“others listen to - has a lot of influence). Nominations were compiled into directed matrices for each classroom. Participants also nominated peers who “hang out together a lot” (Cairns et al., 1985); these were compiled into a co-occurrence matrix for each classroom to assess the “hanging-out” network for each classroom. In preliminary QAP analyses with 3 classrooms, (a) “sighs” was significantly associated (p<.05) with social control but not with influence; (b) “stares” was tied with social control in 2/3 of classes and 2/3 for influence; (c) “keeps out” was associated with social control in 2/3 classes and with influence in 2/3; (d) “stop liking” was associated with social control in 3/3 classes but not with influence (0/3). In a second preliminary step, we summed nominations of influence and social control into sociometric measures of network “reputation”. Moran-Geary statistics indicated influence was nearly unrelated to children’s nominations of specific peers for “sighs”, “stares”, “stop liking”, and “keeps out”. In contrast, a network reputation for social control was related to more arguably detectable forms of social aggression (“stares”, threatens to “stop liking”). Next steps include examining the effectiveness of these forms of social aggression at the subcomponent versus whole network level of the “hanging-out” network and as moderated by network centrality. Metafor will be used to conduct meta-analyses of effects across the 21 classrooms.



← Schedule