Reliability and validity of a demand curve measure of alcohol reinforcement.

Recent clinical research suggests that several self-report behavioral economic measures of relative reinforcing efficacy (RRE) may show utility as indices of substance abuse problem severity. The goal of the present study was to evaluate the reliability and validity of the Alcohol Purchase Task (APT), a RRE measure that uses hypothetical choices regarding alcohol purchases at varying prices (demand curves) to generate several indices of alcohol-related reinforcement. Participants were 38 college students who reported recent alcohol consumption. Both the raw alcohol purchase/consumption values and several of the computed reinforcement parameters (intensity & Omax) showed good to excellent 2-week test-retest reliability. Reinforcement parameters derived from both a linear-elasticity (Hursh, Raslear, Bauman, & Black, 1989) and an exponential (Hursh & Silberberg, 2008) demand curve equation were generally less reliable, despite the fact that both equations provided a good fit to participants' reported consumption data. The APT measures of demand intensity (number of drinks consumed when price = 0), Omax (maximum expenditure), and elasticity (alpha) were correlated with weekly drinking, alcohol-related problems, and other self-report RRE measures (relative discretionary monetary expenditures toward alcohol and/or relative substance-related activity participation and enjoyment). Demand intensity was uniquely associated with problem drinking in a regression model that controlled for weekly consumption. These results provide support for the reliability and validity of the RRE indices generated with the APT.