Classroom gender composition and children's peer relations.

Gender segregation in the classroom is advocated as academically beneficial, particularly for girls. However, the social impact for children has received little attention. The present study compared children's peer relations following the transition from mixed-sex fourth-grade classrooms to same-sex fifth-grade classrooms, and beyond into same-sex sixth-grade classrooms. Participants completed self and peer measures of social competence. The change to same-sex classrooms was associated with more mutual friendship nominations among boys in both fifth- and sixth-grades. Peer nominated overt and relational aggression, victimization, rejection, and passive/withdrawn behaviors increased for girls in fifth-grade, but decreased in sixth-grade. Implications for children's social development are discussed in view of research showing academic benefits of same-sex classrooms.

**********

Gender of peers is a powerful influence on children's social lives. Indeed, during middle childhood, children appear to operate within separate cultures defined by gender (Maccoby, 1995). Certainly, given the importance of the school context for children's development (Minuchin & Shapiro, 1983), the gender composition of a classroom will influence children's social interactions, relationships, and peer social competence. Although same-sex classroom composition has received much attention as an academic issue (e.g., Carpenter & Hayden, 1987; Mael, 1998; Riordan, 1985; Singh, Vaught, & Mitchell, 1998; Vezeau, Bouffard, & Chouinard, 2000; Warrington & Younger, 2001), little research has examined the relation between same-sex classroom composition and children's peer relations. The present research took advantage of the opportunity to assess peer relations in an elementary school, following a move from mixed-sex fourth grade to same-sex fifth classrooms, and beyond into same-sex sixth grade classrooms. Social competence of children in terms of both self-perceptions and peer relations were assessed. As introduction, we present research on gender and children's social development, groups and gender, and the gender composition of the classroom.

Gender and Social Development

Peer aggression and victimization. A clear sex difference often is reported in the form of aggression used by boys versus girls. Aggression is likely to be evaluated negatively by many children (Rys & Bear, 1997). However, boys often employ overt aggression, and girls employ relational aggression (Crick & Grotpeter, 1995; Owens & MacMullin, 1995). Overt aggression occurs in the form of verbal or physical threats and intimidation, and relational aggression involves intentional manipulation and damage of peer relationships (Crick & Grotpeter, 1995).

Gender differences in forms of aggression are likely linked to the different behavioral styles and friendship networks of boys and girls. Boys' use of overt aggression is consistent with a more direct behavior style. Girls' use of relational aggression is perhaps linked to the importance given to relationships and social groups among girls (Owens & MacMullin, 1995). Relational aggression may be most effective in a tight friendship network unlike the larger peer networks more typical of boys' in which social boundaries may not be as clearly defined (Lagerspetz & Bjorkqvist, 1994). Both overt and relational forms of aggression were assessed in the present research. Knowledge of the relation between gender and forms of aggression is based exclusively on data from mixed-sex classrooms, and has not been examined in same-sex environments. The present study offers a rare chance to examine relations between children's gender and types of aggression in the context of only same-sex peers.

Less is known about links between gender and victimization by aggressive peers. Peer victimization often is associated with perceptions of poor competence and low global self-worth in middle childhood (Callaghan & Joseph, 1995; Neary & Joseph, 1994). Frequent and enduring victimization by peers is positively related to loneliness in children (Kochenderfer-Ladd & Ladd, 2001). Behavioral correlates of victimization in childhood included submissive behaviors and reactive aggression (Schwartz, Dodge, Coie, Hubbard, Cillessen, Lemerise, & Bateman, 1998).