An alcohol quantity-frequency composite was calculated at each wave (for this analyses, corresponding assessments to the impulsivity measure were utilized at approximately ages 18, 25, 29, and 35) by averaging two measures of alcohol consumption: (i) total quantity/frequency (QF) of alcohol consumption, calculated from individual QF estimates for beer, wine, wine coolers, and hard liquor during the past 30 days; (ii) QF of alcohol consumption per week based on the past year. Both past-month and past-year estimates were utilized because the former are less likely to be influenced by recall bias, whereas the latter are less likely to be affected by ephemeral variability in drinking patterns.
