PREDICTING THE LONGITUDINAL COURSE ASSOCIATED WITH AGGRESSIVE-REJECTED, AGGRESSIVE (NONREJECTED), AND REJECTED (NONAGGRESSIVE) STATUS

.A 2-year follow-up was undertaken of boys identified in grade school as aggressive-rejected, aggressive (nonrejected), rejected (nonaggressive) or nonproblematic comparison. When the social outcomes of boys in each group were compared, aggressive-rejected children exhibited more conduct problems and social rejection at Time 2 than boys in the other aggressive or rejected groups. Behavior observations, teacher ratings, peer ratings, and open-ended peer interviews collected at Time 1 were used to test three predictive models that might account for the greater developmental risk of aggressive-rejected boys. Multiple regression analyses suggested that a broad array of conduct problems, low levels of positive interaction skills, and high levels of peer dislike and ostracism each made unique contributions to the stability of social adjustment problems of boys in the sample. Within group analyses of change suggested that prediction models varied for boys in the three risk groups. The implications of findings are discussed for developmental models of social risk and for intervention design.