Figure 1.

The standardized effects of the path model. Although all variables were entered into the same model, we split the figure by substance use outcome for clarity. The covariances among distal antecedents, between alcohol and marijuana protective behavioral strategies (PBS), between alcohol and marijuana use quantity, and between alcohol- and marijuana-related consequences, are not depicted for parsimony but are available on request. The direct effects of distal antecedents on alcohol/marijuana consumption and consequences are not shown in the figure for parsimony but are shown in Supplemental Tables 2 and 3. Significant associations are in bold for emphasis and were determined by a 99% bias-corrected standardized bootstrapped confidence interval (based on 10,000 bootstrapped samples) that does not contain zero. *Matched paths shown to be significantly different across substances via a Wald’s test (p < .01) (see Supplemental Figures 1 and 2 for more details).