Relations Between Actual Group Norms, Perceived Peer Behavior, and Bystander Children's Intervention to Bullying.

The goals of the study were (a) to predict children's intervention in bullying situations from class-level norms for intervention, as well as child-level perceptions of the number of peers who would intervene, and (b) to determine whether these predictions held when accounting for children's levels of empathy, prosocial behavior, and callous-unemotional traits. Participants were 751 racially and ethnically diverse fourth- and fifth-grade students (53.8% female) in 43 classes. Participants completed peer nominations about which classmates they perceived would intervene during bullying situations. Empathy and callous-unemotional traits were assessed via self report, whereas prosocial behavior was measured through peer report. Using multilevel modeling, each child's intervention in bullying was positively predicted from class-level norms for intervention (class means for the percentage of children who nominated each child as intervening) but negatively predicted from child-level perceptions of the number of peers who would intervene, after accounting for the 3 child traits. Class-level findings support past research on group norms which suggest that children are more likely to display a behavior if their peers display the same behavior. Child-level findings support the presence of the "bystander effect" in children's bullying episodes, in which children are less likely to intervene if they believe that more peers will do so. Thus, although children were more likely to intervene in classrooms with cultures that made intervention more normative, within the context of each class's culture, children were more likely to intervene if they perceived that fewer peers would do so.