Following Brown’s (1989) method, interviewers also rated whether each SLE was, “clearly independent,” “probably independent,” “probably dependent,” or “clearly dependent” on the participant’s behavior. Interrater and test–retest reliabilities for these ratings were rs = 0.89 and rs = 0.77, respectively. For some analyses, we grouped these into independent and dependent categories. Independent SLE (iSLE) and dependent SLE (dSLE) were coded as dichotomous variables to indicate whether a participant had experienced at least 1 such event in the past year. Further description on LTCT and event dependence ratings are available elsewhere and not elaborated here (e.g., Kendler and Prescott, 2006; Kendler et al.,1998,1999).