Stress, emotion regulation, and social competence in high-risk children: Testing moderator and mediator models via structural equation modeling

This study examined the effects of stress and emotion regulation on children's social competence. One-hundred and four children (52% girls, 89% African American) were recruited via a residential summer-camp and also through an after-school program. The children interviewed were by their social address considered at-risk (all experienced some combination of incarcerated parent, public housing, and/or low income status). We interviewed both the children and their mentors and teachers. The Stress Index was used to assess the number of stressors the children had experienced within the last year. This sample of children experienced a very high number of stressors in the previous year (mean 6.5 stressors). Children were administered sub-scales of the Early Adolescence Temperament Scale-Revised (EATQ-R, Frustration and Aggression) and two sub-scales of the Matson Evaluation of Social Skills with Youngsters (MESSY, Inappropriate Assertiveness and Appropriate Social Skills). Adults were asked to rate children using the Emotion Regulation Checklist (Negative Lability and Emotion Regulation) and the adult version of the MESSY with the same two sub-scales. Structural Equation Modeling was used to test mediator and moderator effects of stress and emotion regulation in two separate models (adult and child). Results suggested that in the adult model, negative lability mediated the relationship between stress and inappropriate assertiveness and that in the child model, frustration mediated the relationship between stress and appropriate social skills. In the adult model, the emotion regulation sub-scale was found to predict variance in both sub-scales assessing social behavior, but did not mediate the effect of stress. Likewise, in the child model, aggression predicted variance in both sub-scales of social behaviors, but did not mediate the effect of stress. Several cross-measure/cross-informant correlations were significant, suggesting a reasonable agreement between children and adults. Findings indicate that high-risk children who are less socially skilled may benefit more from guidance that focuses at reducing negative social behavior rather than teaching them specific pro-social skills.