Reliability of Self-Reported Economic Data From MDMA Users in St. Louis, Miami, and Sydney

Abdallah, Arbi Ben; Copeland, Jan; Inciardi, James, A.; Cottler, Linda, A. Washington University/ University of New South Wales, Sydney, Australia/ University of Delaware, Newark, DE, U.S.A./ Washington University, St. Louis, MO, U.S.A.

Though there has been much interest in cross-cultural studies of the consequences of substance use, less interest has been focused on the economic data provided by substance users. Further, there liability of this self-reported economic data is generally nonexistent. Previous studies suggest substance users withhold information during interviews, leading to response or measurement errors. Although numerous factors (associated with the interview questionnaire, the interviewer, and interviewee) could be associated with measurement errors in self-reporting, much of the empirical evidence available about the quality of self-reported information by substance users has focused on recall bias. Although most substance users want to provide the best information they can, they may not understand the questions, or have difficulty accurately recalling past events. Moreover, the use of drugs in general, and Ecstasy in particular, has been shown to affect memory and cognitive skills. Reliability of self-reported data in general, and economic data in particular, is even more important in the presence of cultural differences and a growing interest in the economic aspects of drug abuse. Using data from our NIDA-funded “Tri-City Study” of the reliability of DSM-IV abuse and dependence criteria, the present research had the opportunity to assess the reliability and feasibility of collecting economic information from MDMA users in three geographically separate regions (St. Louis, Miami, and Sydney). Recreational MDMA users (Total N =590), (St. Louisn1=270, Miami n2=180, Sydney n3=140) were recruited and interviewed twice, one week apart, with the same survey questionnaires. Proper reliability methodologies were used, including independent observers for the time-1 and time-2interviews. Item non-response was used as a measure of acceptability. As a test-retest reliability measure, Cohen’s Kappa (K) statistic for categorical responses and Intraclass Correlation Coefficient (ICC) for continuous data were applied. Ecstasy users were found to be not only willing to provide information on their employment, earnings and drug cost details, such as frequency and number of Ecstasy pills used, as well as cost and time spent acquiring, using, and being high, but were also consistent in reporting their information (K ranged between 63 and 85 and ICC ranged between 54 and 99). Differences by site are discussed and used to draw specific implications for future cross-cultural research on MDMA use.