Friendship and Aggression in Elementary School. The friendships of aggressive children and the effects of having aggressive friends

Research has demonstrated the importance of friendship for children’s adjustment (e.g., Ladd, 1990). However, some children may be less capable of maintaining satisfying relationships with their peers. Aggressive children have been found to experience peer relation difficulties (e.g., Dodge, Coie, Pettit, & Price, 1990). Most research on aggression and peer relations has focused on children’s status within the peer group (group interaction), rather than friendship (dyadic interaction). Group interactions and dyadic interactions are unique aspects of social experience, though, and therefore it is important to differentiate between them in research (Erdley, Nangle, Newman, & Carpenter, 2001). Friendships may be more intimate, and therefore may have more impact on children’s development. This thesis’ first aim was to examine the dyadic friendships of aggressive children. We demonstrated that aggressive children did not experience difficulties with their friendships, but were able to form mutual friendships of good quality. We even identified a subgroup of aggressive children (Machiavellians) whose friendship experiences resemble those of prosocial children. Secondly, we aimed to examine the adjustment of children who befriend aggressive children. Research in adolescence has shown that involvement with deviant peers can have a negative influence on children’s own adjustment (e.g., Patterson, Reid, & Dishion, 1992). However, this area of research mainly focuses on adolescents who come from families with low socio-economical status, and adolescents who display severe forms of antisocial behavior or risk behaviors that specifically gains importance in adolescence (e.g., smoking). It is important to examine whether or not the same patterns of peer influence also hold for younger children in regular elementary schools. At this age, children show other forms of deviant behavior that can be precursors for more severe antisocial behavior in adolescence. We found that child characteristics, and to a lesser extend friend characteristics, contribute to the concurrent prediction of adjustment. Remarkably, having aggressive friends did not seem to play a role in predicting internalizing and externalizing problems, when controlled for children’s own levels of aggression. In contrast, friend active isolation and passive withdrawal did influence a child’s externalizing behavior in specific cases. Further, we found differences in levels of competence and problem behavior for children without friends, children with aggressive friends, and children with nonaggressive friends. Children without friends in grade 1 were less liked over time by their peers and subsequently were lonelier than children with (aggressive or nonaggressive) friends. Furthermore, children without friends and children with aggressive friends in grade 1 displayed higher levels of aggression over time than children with nonaggressive friends. No differences in academic competence were found between groups of children. Although main effects appeared, we did not find any of the hypothesized moderating effects of friendship on the longitudinal relations between competence and problem behavior. Thus, the studies in this thesis cross-sectionally and longitudinally showed that having aggressive friends in elementary school did not augment children’s maladjustment over time (nor did having nonaggressive friends buffer against maladjustment). However, children without friends in the beginning of elementary school clearly showed signs of maladjustment over time