Figure 3 summarizes the results of the path analysis, and results are shown in Table 4. Contrary to our hypothesis, EXT PGS was not associated with externalizing behavior, parental knowledge, and peer substance use at T1. However, consistent with our expectation, all stability paths between T1 and T2 (e.g., T1 parental knowledge and T2 parental knowledge) were significant (ps < .01). Further, within time covariation paths (e.g., T1 parental knowledge with T1 externalizing behavior) were all significant (ps < .01) in the expected direction. Controlling for covariation links among domains within time or stability within domains between T1 and T2, T1 parental knowledge predicted lower T2 externalizing behavior (p < .05). T1 peer substance use was, however, not significantly associated with T2 externalizing behavior. T1 externalizing behavior predicted lower levels of parental knowledge at T2 and higher levels of peer substance use at T2. There was no evidence of indirect effects of EXT PGS on T2 externalizing behavior via any of the T1 variables.
