Following these preliminary analyses, we fit our theoretically informed path model in Mplus version 6 (Muthén & Muthén, 1998-2011). As shown in Figure 1, this model tests whether parental alcohol problems have direct and indirect effects on follow-up romantic relationship status, and whether romantic relationship status is part of an indirect pathway linking these alcohol risk factors and alcohol use. In addition to these paths of primary interest, we also statistically controlled for the direct effects of all variables and covariates on alcohol use. This allowed us to examine whether our hypothesized indirect paths of interest were significant above and beyond the direct paths from parental alcohol problems, conduct problems, and positive urgency on alcohol use, as well as the indirect paths from parental alcohol problems→conduct problems→alcohol use and from parental alcohol problems→positive urgency→ alcohol use.
