Mediation of Treatment Effects in Problem-Focused Social Skills Training versus Resource Activation for Children with Disruptive Behavior Problems: A Cross-Sectional Analysis

A previous randomized controlled trial comparing problem-focused computer-assisted skills training (ScouT) with supportive resource activation treatment (STARK) in children (6-12 years) with disruptive behavior disorders (DBD) demonstrated significant pre-post changes in both treatment conditions, with ScouT being slightly superior on some outcomes. Based on the same sample (N = 99), the present study examined cross-sectional indirect associations between treatment condition and outcomes via intermediate variables, which conceptually correspond to mediation. We hypothesized that ScouT would lead to stronger reductions in the peer-related aggression-maintaining factor (PAM factor; including dysfunctional social information processing, emotional dysregulation, impulse control problems, deficits in social skills, and dysfunctional social interactions), whereas STARK would lead to greater improvements in self-concept, which in turn would be associated with lower concurrently assessed clinician-, parent- and teacher-rated DBD symptoms and overall externalizing behaviors. Our analysis yielded support for the hypothesized indirect association through the PAM factor, with consistency across all outcome measures. Unexpectedly, compared with STARK, ScouT also led to a more favorable self-concept, which was indirectly associated with less severe clinician-rated DBD symptoms, but not with the other outcomes. This latter indirect association was small and should additionally be interpreted with caution given the low reliability of the clinician-rated scale. These findings may indicate that the treatment of children with DBD may benefit from directly targeting aggression-maintaining factors. Limitations include the concurrent assessment of the intermediate variables (PAM factor, self-concept) and the outcomes, which does not permit causal conclusions. Future research should refine methodologies and explore additional treatment mechanisms.Trial Registration: The study was registered at clinicaltrials.gov on 2014-05-18 (Identifier: NCT02143427).