Antisocial and aggressive behavior among children and youth is a nationally recognized public health problem. From a social learning theory perspective, research suggests that early aggressive behavior is learned through negative reinforcement mechanisms underlying a coercive family process in which the principal socialization agent is the parent. It is unclear what specific social processes occur and come to regulate aggressive behavior when the child enters the preschool environment. Presumably, repeated displays of aggressive acts affect the way adults (i.e., teachers) and peers respond to and interact with the aggressive child. However, the relative effects of teacher and peer interactions in preschool environments are not well understood. A sample (N=72) of preschool children characterized as high aggressors will be randomly selected from three urban alternative schools serving young children with severe behavior problems. Using a repeated measures longitudinal design, standardized direct-observation data will be collected monthly for all subjects. Within-subject lag sequential analysis will be used to determine the magnitude of association between aggressive behavior and social consequences on a session-by-session basis. A dynamic covariate analysis will be used to test the effects of the victim's response (i.e., reciprocated aggression), the sex of the victim, and the sex of the aggressor on changes in teacher responsivity. To test the effects of changes in teacher responsivity on changes in the base rate of aggression over time, longitudinal hierarchical linear modeling (HLM) will be used. Longitudinal studies are needed to determine how early patterns of high-rate aggression in preschool classrooms is influenced by social context.
