Despite these strengths, our findings need to be interpreted in view of several limitations. First, our analyses focused on a sample of European ancestry ascertained primarily from large extended families enriched for alcohol use disorders. Thus, generalizability of our findings to samples of non-European ancestry and community samples is unclear and additional research is needed to replicate our findings. Second, we only examined one aspect of positive parenting (i.e., parental bonding) due to the limited parenting measure available in COGA with good reliability and validity. Future studies should consider the role of other dimensions of parenting and home environment and across different developmental periods that may be more explicitly relevant and salient for promoting educational achievement (e.g., parents reading to offspring and helping with schoolwork), as well as other psychosocial and environmental pathways of the social genetic effects. In addition, our measure of parenting was based on offspring-report only and may be subject to bias. Future research is needed to replicate our findings using other methods to measure parenting behaviors (e.g., parent-report or observation).
