In addition to these hypothesized indirect effects via dating several people, we also statistically controlled for and examined additional direct and indirect effects. Conduct problems and positive urgency had direct effects on alcohol use (p ≤ 0.01), but parental alcohol problems did not (p = 0.85). There was a significant indirect effect from parental alcohol problems→conduct problems→alcohol use (indirect effect: B = 0.02, p < 0.01, 95% CI [0.005, 0.028]) and from parental alcohol problems→positive urgency →alcohol use (indirect effect: B = 0.008, p = 0.02, 95% CI [0.002, 0.015]). In an exploratory analysis we examined potential sex differences using an omnibus test of whether the path coefficients could be equated across males and females. The result of our omnibus test of sex differences was not significant (Wald test of parameter constraints: χ2 (10) = 12.78, p = 0.24). Thus, there did not appear to be sex differences in the effects.
