Predictors identified of engagement in suspendable school behavior include: (1) observation of violence at school, suggesting girls may be learning to behave violently by observing violence in school; (2) gang involvement; and (3) a favorable attitude toward violence. The girls in the study also appeared to be making distinctions not between types of aggression but between violence used for aggression and violence used for the sake of defense. Two other intervenable social variables, prior school suspension and drug use, were more weakly associated with suspendable school offenses. The purpose of this study was to determine the relative strength of predictors of school violence among a sample of 229 girls enrolled in a middle school. References