This assessment proceeded in two steps: For the first five situations, adolescents provided their responses without any indication as to how others would respond. For the second five situations, adolescents were first told of a hypothetical peer's deviant response (e.g., “One teen stated that they would take the sweater from the store and run.”). The degree to which adolescents' rated skill levels declined from the first five situations to the second five situations (e.g., after hearing another teen's deviant response) was used as a marker of presence or absence of skills in the face of deviant peer behavior. This measure was obtained repeatedly at each of the first 3 waves of data collection (ages 13, 14, and 15) using different items at each wave, and the mean of scores across three waves was used in analyses. Overall reliability for this final measure, calculated using the intraclass correlation coefficient, was r =.87, which is considered in the “excellent” range for this statistic (Cicchetti & Sparrow, 1981).
