Epidemiology and Etiology: Testing the Cross-Lagged Relationships between Parental Monitoring and Antisocial Peer Associations in a Longitudinal Adolescent Sample

This abstract was presented at the 2018 Society for Prevention Research Annual Meeting which was held May 29 – June 1, 2018 in Washington, DC, US.

Christopher Mehus University of Minnesota-Twin Cities

Myriam Forster California State University, Northridge; Gary C. K. ChanUniversity of Queensland; Sheryl A. Hemphill Australian Catholic University; John Winston Toumbourou Deakin University; Barbara McMorris University of Minnesota

Introduction: Parental monitoring and antisocial peer associations are both risk factors for negative behavioral outcomes. Previous research has yielded somewhat disparate findings regarding the causal relationships between parental monitoring and antisocial peer association or problem behavior. Few studies have collected the necessary longitudinal data to examine these relationships in a sample followed throughout adolescence. Additionally, few studies have measured antisocial peer associations rather than problem behavior. The temporal order of these variables has significant implications for family-focused preventive interventions. We tested these reciprocal relationships across four waves of an adolescent sample. Based on the Social Interaction Learning model, we hypothesized that poor parental monitoring would predict increased association with antisocial peers but that the reverse would not be significant. 

Methods: We used four waves of data (grades 5, 7, 9, 11) from students in Victoria, Australia (n=922) participating in the International Youth Development Study (IYDS). The IYDS used a modified version of the Communities that Care survey. We used MPlus 7.2 to construct the cross-lagged path model, controlling for gender. Poor monitoring was a six-item scale of youths’ self-reported answers to items that asked about parental monitoring and parent’s knowledge of their activities. Anti-social peer associations was a six-item scale of items related to illegal or delinquent behavior of the youth’s friends in the last year. 

Results: The specified reciprocal model fit the data well (χ2(12) = 56.141, p < .001; RMSEA = .06 [90% CI: .047 - .080]; CFI = .96; SRMR = .04). Across all data waves, poor parental monitoring predicted higher antisocial peer associations at the next wave, whereas antisocial peer associations only predicted poorer monitoring from the 7th to 9th grade waves. 

Conclusion: Findings largely support the temporal ordering proposed by the Social Interaction Learning model. The results will be presented in the context of existing literature in this area, which is laden with complexity related to the operationalization of monitoring and differences related to developmental stages. The results of this study support the predominant view that parental monitoring is a useful target for family-focused preventive interventions.