Our results should be interpreted within the context of the following limitations. First, COGA is a high-risk sample with most participants from extended families enriched for alcohol use disorders. Findings from this study may not generalize to samples with different recruitment/ascertainment strategies (Savage et al. 2018). Replicating our findings in community and population-based samples is an important next step. Second, all of our measures were based on self-report data. It is possible that adolescents’ externalizing traits or their genetic predispositions led to biased reports of perceived peer substance use (Bauman and Ennett 1996) and parental knowledge. Third, our measure of parental knowledge tapped adolescent perceptions rather than actual parental knowledge or specifically implemented parental monitoring behavior/strategies and control (Kerr et al. 2010; Stattin and Kerr 2000). Future work is needed to replicate our findings and to use multi-informant data (e.g., adolescent report, parent report) and consider other dimensions of parenting. In addition, our measure of parental knowledge was not specific to maternal or paternal knowledge. Overall, findings of the current study appear to generalize to males and females. However, it
