Substance use and educational outcomes become associated during adolescence (Townsend et al., 2007) but reasons for the association remain poorly understood. Theoretically, the association could be due to several mechanisms (see, e.g., Haller et al., 2010) which can be either causal or non-causal. First, there could be a causal influence of the educational domain on substance use. Thus, difficulties and poor performance in school might, via mediating psychological and social factors, causally lead to increased substance use. Alternatively, the causal relationship could run in the opposite direction so that substance use would influence educational outcomes. Such an adverse effect could be related to, e.g., cognitive or motivational factors influencing school performance, or substance use could foster contact with peer groups unfavorable for school achievement. Third, the causal relationship could be reciprocal, whereby poor school outcomes and substance use behaviors would mutually influence each other and lead to a self-reinforcing process. Finally, the association could be non-causal and merely reflect confounding factors influencing both educational and substance use outcomes. For example, familial background may be related both to education and substance
