The processes by which children are classified as aggressive have important educational and research implications. For example, aggression in childhood reliably predicts dropping out of school and incarceration. The author argues that the sex-role stereotypicality of aggression produces bias in both observers and raters of student behavior. The tendency is to overattribute aggression to boys, relative to girls, resulting in questionable validity of assessments. The author reviews these issues, proffers some explanations (e.g., gender schema theory), and makes policy recommendations (e.g., foregrounding the importance of social competence as an educational outcome). (Contains 1 note.)