The ALSPAC dataset offers many important attributes, and this is a valuable population in which to measure aspects of the hypothesized LR-based model. However, as is true of all studies, it is important to recognize potential caveats in interpreting the results. First, the population, while large, was selected from only one section in the UK, and was mostly Anglo-European in background. As a non-clinical sample it is likely that the impact of extreme psychopathology or extensive substance use in the offspring and their parents could not be adequately evaluated, and these factors were not considered in the analyses. Second, the questionnaires used to measure the domains were relatively limited in scope, and it is possible that different findings might have resulted if alternative measures had been used. For example, the score for drinking in peers represents the offsprings’ perceptions of what they believed their peers drank and did not include a potentially more accurate assessment of the peers themselves. In addition, the SRE only broadly defines the time period of interest up to the First 5 times of drinking, with
