Forty-five subclinical families with behavior-problem children will serve as subjects in a study of family intervention. Subjects will be screened to eliminate all parents and children with a recent history of treatment for psychiatric or psychological disorders. Two strategies--parent-training-plus-child-training versus parent-training alone--will be compared with that of a waiting-list control group. The effects will be assessed on changes in (a) the parents' and child's positive and negative behavior directly observed in the home environment by trained observers, (b) the child's positive and negative behavior recorded daily by parents, and (c) the daily ratings of parent and child satisfaction with each other. These measures will be collected before, during, and after treatment. A pre-post, between-group, matched-subjects experimental design will be used. Families will be matched on variables previously shown to affect severity of child behavior and parent training outcomes. These are marital satisfaction, maternal depression, and socioeconomic status. Conditional probability analyses will also be used to examine between-group differences in reinforcement and punishment behavior, before, during, and after treatment. In the parent-training condition, parents will receive small group parent training employing didactic instruction, behavioral rehearsal, and therapist feedback. In the parent-training-plus-child-training condition, parents will receive parent-training identical to that in the parent-training condition. However, the target children will receive concurrent social skills training using a direct instruction format. These prosocial skills are intended to increase the child's positive interactions with parents. Examples of skills taught include showing appreciation and conversational skills. Follow-up data at six months and one year will determine the relative effectiveness of the two strategies in maintaining change.
