The Effects of Neighborhood Safety and Deviant Peer Affiliation on Child Aggression: A Multi-Informant Analysis with Emotion Regulation As a Mediator

Background and Purpose: Childhood aggression can have short-term and long-term negative impacts on children's overall development, such as poor school adjustment, mental health problems, and offending behaviors in adulthood. Neighborhood safety and deviant peer affiliation are important social-environmental determinants of childhood aggression. Although studies have examined the direct pathways from neighborhood safety and deviant peer affiliation to child aggression, the psychosocial mechanisms underlying these pathways are largely unknown. Literature suggests that internal characteristics, such as emotion regulation, could be the potential mediator. In addition, most existing studies employed data from one informant (eg., solely parent-report data), which may lead to inflated estimations in relationships. Therefore, according to the neighborhood disorder model and peer contagion theory, this study used multi-informants (parents, teachers, and children) to examine the direct effects of neighborhood safety and deviant peer affiliation on child aggression, as well as the indirect effects of emotion regulation.

Methods: Data were derived from the elementary school success profile (ESSP), where 3^rd to 5^th graders were selected from 13 elementary schools in four districts in North Carolina during the 2008-2009 academic year. A total of 682 students who had information from their caregivers and their teachers were included in this study. This study employed a multi-informant approach: information about neighborhood safety and child emotion regulation was collected from parents, information about deviant peer affiliation was collected from the child, and information about child aggression was collected from teachers. Mplus 8.0 was used to examine the proposed theoretical model. Structural equation modeling (SEM) with weighted least squares with mean and variance adjusted (WLSMV) estimator was conducted using the bootstrapping approach (n = 2,000).

Results: The SEM analysis showed a good fit of the model to the data (CFI = 0.992, TLI = 0.991, SRMR = 0.049, and RMSEA = 0.046). While neighborhood safety was not directly associated with child aggression, deviant peer affiliation was directly related to child aggression (β = 0.225, p < 0.001), controlling the effect of gender on child aggression (β = -0.233, p < 0.001). The indirect effects of neighborhood safety and deviant peer affiliation on child aggression through child emotion regulation were, respectively, -0.085 (SE = 0.020, CI = [-0.140, -0.052], p < 0.001) and 0.055 (SE = 0.017, CI = [0.026, 0.094], p < 0.010). The direct and indirect effects of gender, neighborhood safety, deviant peer affiliation, and child emotion regulation accounted for 25.8% of the explained variance in child aggression.

Conclusions and Implications: This study empirically supported the proposed model of how social-environmental factors (i.e., neighborhood safety and deviant peer affiliation) influence child aggression through the children's internal characteristics (emotion regulation). Hence, policymakers who aim to reduce child aggressive behaviors may consider developing prevention and intervention programs to improve neighborhood safety. Caregivers and school social workers may consider minimizing children's interaction with deviant peers and substituting them with more constructive alternative activities. In addition, strategies for enhancing children's emotion regulation abilities are warranted for those affiliated with deviant peers and living in neighborhoods with low safety.