Existing cross-lagged analyses of educational achievement and drinking or smoking behaviors have been methodologically variable and have produced differing results. Thus, among nearly 12,000 middle and high school students from the National Longitudinal Study of Adolescent Health, academic failure predicted alcohol use (but not binge drinking) one year later, but only a suggestive effect of alcohol use on academic failure was noted (Crosnoe, 2006). In contrast, a much smaller study of 405 participants in a high-risk sample found no cross-lagged associations between academic achievement and binge drinking across three study waves spanning an 18-year period (Haller et al., 2010). Bidirectional cross-lagged associations between school achievement and smoking across adolescence were indicated in two large studies (Pennanen et al., 2011; Tucker et al., 2008). In contrast, in data from the Monitoring the Future project academic achievement negatively predicted smoking from 8th to 12th grade but the reverse was not found (Bryant et al., 2000).
