Overall, the model demonstrated good fit across multiple indicators (X2 = 47.92, p < 001; RMSEA =.05; CFI =.98; TLI =.83). Figure 1 displays the model graphically and Table 2 displays the unstandardized and standardized beta coefficients for direct and indirect paths among the study variables for girls. Direct paths between maternal problem drinking and adolescent drug use (β = .15, p < .01), rule breaking (β = .40, p < .001), and aggressive behavior (β = .25, p < .05) were observed for girls. There were no significant direct paths between paternal problem drinking and externalizing behaviors. Indirect paths showed that both maternal (β = -.09, p < .05) and paternal (β = -.20, p < .001) problem drinking negatively predicted family cohesion. Family cohesion, in turn, was negatively related to both adolescent rule breaking (β = -.47, p < .001) and aggressive behavior (β = -.44, p < .01). These findings suggest that there is an indirect effect between parental problem drinking and later adolescent externalizing behaviors through family cohesion. Family cohesion did not predict any of the substance use outcomes.
