Interestingly, results from our multivariate SEM model indicated that fathers’ but not mothers’ positive parenting was associated with fewer adolescent conduct problems, whereas mothers’ but not fathers’ positive parenting was associated with adolescents’ lower likelihood of engaging in risky drinking. This is unexpected, given that both paternal and maternal parenting behaviors have been shown to be associated with risky drinking and externalizing problems in adolescents (Leidy et al., 2011; Ryan et al., 2010). In our study, when paternal and maternal parenting behaviors were analyzed in separate models (post-hoc analysis, see supplemental Figures 1 and 2), both were indeed significantly associated with adolescent risky drinking and conduct problems, as also indicated by the bivariate correlations between these variables. Our model where fathers’ and mothers’ parenting analyzed simultaneously took into account the variances shared between fathers and mothers and provided an examination of the unique effects of fathers’ and mothers’ parenting. It is interesting that results from the fathers and mothers combined model and those from the analyses conducted separately for fathers and mothers were somewhat different. This suggests that while some
