Social attributional biases of peer-rejected and aggressive children.

Peer-rejected high-aggressive, rejected low-aggressive, and non-rejected third- and fifth-grade boys were shown sets of drawn pictures differently balanced on Kelley's social dimensions of distinctiveness (i.e., depicting a hypothetical peer interacting with other children) and consistency (i.e., depicting the peer interacting with the subject) information. Following each set, a provoking incident involving the subject and peer was described in which the subject experienced a negative outcome and the peer exhibited ambiguous intent. When given no social information, rejected high-aggressive and rejected low-aggressive boys made more hostile attributions and suggested more hostile responses. When provided social information, however, all groups made similar attributions and weighted consistency information more heavily in their evaluations. All groups maintained a consistent response style across information conditions, with rejected high-aggressive boys suggesting the most aggressive responses. Implications for the treatment of childhood aggression are discussed.