A noteworthy finding in Dawson and colleagues’ (2007) paper is that when analyses were restricted to a set of 4 SLE deemed unlikely to be consequences of drinking (“death of a family member or close friend, illness of a family member or close friend, change in living situation, and change in work hours or responsibilities”; p. 71), the effect size for the interaction between SLE and drinking onset was similar to that obtained in the analyses for the full set of 12 events. This suggests that the impact of the independent events was similar to that of the other SLEs. We tested this explicitly using measures of event dependence coded from interviewer ratings rather by inferring independence based on type of event. If early drinkers have greater consumption when faced with stress and behave in ways that lead to more SLE, we would expect to see a significant onset × SLE interaction for both dependent and independent events.
