Differential Benefits of Skills Training with Antisocial Youth Based on Group Composition: A Meta-Analytic Investigation. (Special Topic)

Article excerpt


A meta-analysis of 38 studies of social skills training interventions with antisocial youth was performed. Also examined were treatment effects for interventions that differed in group composition. The intervention studies yielded an overall effect size of .62 at posttreatment. As predicted, skills training interventions delivered in the context of groups consisting of only antisocial peers produced smaller benefits than did skills training interventions that avoided aggregating antisocial peers (i.e., groups comprised prosocial and antisocial youth or individual treatment). For those 18 studies for which follow-up data were reported, treatments provided in the context of either mixed or individual treatment also produced larger follow-up effect sizes than did deviant-only group interventions. These findings add to a growing body of evidence of smaller treatment benefits associated with skills training provided in the context of groups composed exclusively of antisocial participants. Implications for the recr uitment of prosocial peers for inclusion in skills training interventions for aggressive and antisocial youth are discussed.


Research on interventions designed to reduce childhood aggression and prevent adolescent delinquency and substance abuse has expanded rapidly in the past decade. An increased understanding of the causes and consequences of antisocial behavior in children and adolescents and the rapid rise in the rate of juvenile crimes in the 1980s and early 1990s stimulated this research. One of the most popular treatments for youth with conduct problems in schools and clinics is group-based skills training, in which youth are taught a set of social or problem-solving skills to help them better negotiate problem situations without using aggressive means (Kazdin, 1997). Group-based skills training offers an economical and convenient means of providing psychological services to youth with conduct problems. Whereas some group-based interventions have produced positive outcomes, others have failed to do so, and some interventions result in detrimental outcomes (see Arnold & Hughes, 1999).

In a meta-analytic investigation of group social skills training with children experiencing a range of problems, Beelmann, Pfingsten, and Losel (1994) found that group-based skills training resulted in modest short-term gains but that long-term benefits were generally lacking. Also, short-term benefits were greater for measures of targeted skills than for measures of social adjustment. Because Beelmann et al. included studies with children presenting with a range of problems, including antisocial behaviors, their findings may not generalize to studies of skills training with antisocial youth (1). A quantitative summary of the effects of skills training with antisocial youth would assist in evaluating the efficacy of this treatment with this population. In particular, it is important to determine if treatment effectiveness differs as a function of group composition.

In a narrative review of group-based skills training with aggressive youth, Arnold and Hughes (1999) argued that the composition of group participants in skills training programs may affect gains. Specifically, they suggested that the expected benefits of skills training with antisocial youth may be diluted due to unintended negative effects resulting from aggregating antisocial youth. They presented four types of evidence to support their argument that group-based treatment may result in iatrogenic effects that erode the positive gains youth accrue from treatment. First, several developmental studies support the view that affiliation with deviant peers contributes to antisocial developmental trajectories (Curran, Stice, & Chassin, 1997; Dishion, Andrews, & Crosby, 1995; Dishion, Capaldi, Spracklen, & Li, 1995; Dishion, Spracklen, Andrews, & Patterson, 1996; Elliott, Huizinga, & Ageton, 1985; Kandel, 1973; Keenan, Loeber, Zhang, Stouthamer-Loeber, & Van Kammen, 1995). …