Consistent with our hypothesis, results indicated that there were significant indirect effects of EXT PGS on T2 externalizing behavior via T1 variables. Specifically, EXT PGS was associated with higher T2 externalizing behavior via higher levels of peer substance use at T1 (B = .02, 95% CI [.004, .039], p < .05). Further, as expected, EXT PGS was also associated with higher T2 externalizing behavior via more externalizing behaviors at T1 (B = .07, 95% CI [.030, .123], p < .01). There was no significant indirect effect of EXT PGS on T2 externalizing through T1 parental knowledge (B = .01, 95% CI [.000, .032], p = .145).
