Selection of twins for the ERP laboratory study was based on their intrapair differences in RAPI scores (22 items with 4 response alternatives to yield a 66-point potential range) at age 18½. Extremely RAPI-discordant twin pairs were of major interest for their informational yield in planned candidate gene studies and for evaluating possible effects of differential exposure to alcohol during adolescence in within-family comparisons. RAPI scores at age 18½ significantly correlated (0.39) in the full sample of > 2,500 twin pairs, and to no surprise, more highly for MZ than same-sex DZ twin pairs. Accordingly, RAPI-discordant co-twins are disproportionately DZ twin pairs; RAPI-discordant MZ twin pairs are more often twin sisters than twin brothers, and in extremely RAPI-discordant brother–sister twin pairs, it is the brother, more often than the sister, who has the high RAPI score in late adolescence.
