| Teacher-reported, peer-reported, and observed aggression were evaluated in 103 (59 boys, 45 girls) aged 67 to 106 months (M = 85.46, SD = 4.02). Participants included children at preoperational (n = 28), transitional (n = 23), and concrete operational (n = 52) Piagetian developmental levels. Relationships among aggression ratings, as well as the relationships between aggression, developmental level, and social status were examined. Teacher ratings of aggression were found to be significantly related to developmental level, whereas behavioral observations and peer reported aggression were not. In addition, teacher and peer reported aggression were significantly related to each other, but neither were consistent with actual observed aggression. Findings suggest that teachers and peers are taking different factors into account, or are basing their reports on factors that are not being captured by independent observers, when rating aggression. Implications for assessment of aggressive behavior in the schools and for future research are discussed. |