Finally, on a post-hoc basis, we created a composite of the susceptibility measures by standardizing then summing the variables that interacted with peer substance use in the models above (adding social acceptance and teen recantations directly, and reverse-scoring and adding teen social skills and maternal support). As shown in Table 5, this factor, which might be considered as a composite of the risks of susceptibility to peer influence, strongly interacted with peer substance use to predict relative changes in future teen substance use (β = -.33, p < .001) and accounted for 11% of the total variance in adolescent substance use at age 16, even after accounting for baseline use and demographic factors.
