One of the unique features of our sample is that it includes observations in two distinct generations, parents (who were middle-age at the time of assessment) and their offspring (who were between 16 and 21). One might expect that the mean levels of each of the five behavioral disinhibition phenotypes would vary between the parent and offspring generations. Consistent with this expectation, generation had a large covariate effect in the RFGLS analysis. Of greater significance in the present analysis, however, is whether any SNP effects varied across the generations. In order to account for this possibility, we included a SNP by generation interaction term in the regression models and found little evidence for differential effects. Moreover, biometric analysis of the familial data suggested that while collateral first-degree relatives (i.e., biological siblings, dizygotic twins) were consistently more similar than intergenerational relatives (i.e., parents and their offspring), this increased similarity could be attributed to greater environmental similarity among sibling as compared to parent-offspring pairs (i.e., to shared environmental effects). The pattern of familial correlations is consistent with heritable factors having a similar
