There were several purposes of the present study. First, quasi-experimental tasks assessing attributional style were used to investigate the mediating role of social cognitive processes in aggressive behavior for both boys and girls. Secondly, the use of current self-report measures of children's assertive and attributional style allowed for investigation of construct validity of these measures. Third, the relationship of current instruments used in the identification of social dysfunction in children to other measures was investigated. Finally, the use of sociometric ratings by subjects’ peers and teacher's behavioral nominations of subjects for aggressive or assertive status provided information about subjects’ social environment. Fourth and fifth graders identified as aggressive or assertive on the basis of teacher nominations were compared in terms of self-report, other-report, and quasi-experimental data. Results showed sex and status differences on self-report and quasi-experimental measures of attributional style. Selective attention to hostile cues, quick responding, and theoretical ''socially inappropriate" explanatory style are discussed in terms of social cognitive processes involved in aggressive behavior for boys and girls. Further, although current children's assertiveness measures provided good content validity for the behaviors they purport to measure, the conceptual relationship between aggression and social dysfunction was found to be less direct than the literature has indicated. Similarly, self-report attributional style was less directly related to measures of assertiveness and social functioning than was expected. Results are discussed in terms of both content validity of dependent measures and broader theoretical issues.

M.S.