A Mediation Process of Reducing Antisocial Behaviors through Minimizing Deviancy Training

Abundant literature shows that deviant peer associations predict future aggressive behaviors, substance use, and juvenile crimes (Dishion, 1999; Patterson, 1993).  Peer interaction dynamics is a key socializing context for the development of antisocial behavior.  One of the peer interaction dynamics is a process often referred to as ‘deviancy training’ (Dishion, Spracklen et al, 1996) such that youth’s deviant behavior or talk (e.g., being ‘tough’ wins, deviant behavior is funny) is more likely to be responded to by deviant peers with positive affect than behavior or talk that is non deviant. This pattern of interaction has been linked to amplification of adolescent problem behavior in multiple forms. This study is to apply multilevel survival mediation analysis to examine intervention and mediation effects of an intervention project, the Project Alliance 1 (PAL1; Dishion and Kavanagh, 2003).  Specifically we examine 1) the dynamic process of how positive affective responses relate to deviant talk such that longer positive affect would extend the duration of deviant talk bouts, 2) the extent to which deviant friendship process predicts future problem behaviors, and 3) the mediation pathways of the effect of PAL1 in reducing problem behaviors and substance use through its effects on minimizing deviancy training processes.  The PAL1 is an ongoing multiwave longitudinal study of the Family Check-Up (FCU) intervention of improving family and adolescent functions in early adolescence which is hypothesized to reduce later problem behaviors and substance use.  Nine hundred and ninety-eight youth and families (498 control; 500 intervention) were randomly assigned to the FCU program and control conditions. Nine assessments occurred between ages 11 to 28 up to now.

Observation of videotaped adolescent-friend interactions were collected at age 16-17 from 713 youths. Based on observed exchanges between adolescent-peer dyads, ebb and flow of deviant talk and positive affect during the dyadic exchanges were assessed (i.e., multiple deviant topics and affective responses and their duration) and used for the evaluation of deviancy training in the friendship. Antisocial problems at concurrent and latter assessments were the outcome measures.  The preliminary results show that peer positive affective responses reinforced and prolonged the duration (i.e., hazard rate) of deviant talk and such relation was stronger for youth with higher baseline antisocial problems.  Longer deviant talk also positively related to concurrent antisocial problems, controlling for baseline problems. There was a positive intervention effect of reducing deviant talk.  The next step is to examine the mediation effects of intervention to the hazard rates of deviant talk to future outcomes.