Drug and Alcohol Dependence, 9 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG (1982) 273 278 273 zyxwvutsr RELIABILITY OF SELF-REPORTS OF LOW ETHANOL CONSUMPTION BY PROBLEM DRINKERS OVER 18 MONTHS OF FOLLOW-UP STEPHEN A. MAISTO Lkpartment of Psychology, Vanderbilt University, Nashville, Tennessee, 37240 (U.S.A.) MARK B. SOBELL* and LINDA C. SOBELL Clinical Institute, Addiction Research Foundation, 33 Russell St., O ntario M 5S 2Sl and Department of Psychology, University of Toronto, O ntario (Canada) (Received January 25, 1982) Summary Forty-eight male problem drinkers participating in an l&month pro- spective evaluation of an outpatient alcohol treatment program were inter- viewed monthly about their drinking and related behaviors; their collateral informants were interviewed every two months during this same period to corroborate the subjects’ self-reports. A high degree of consistency was found between subjects’ and collaterals’ reports of the subjects’ drinking behavior, including days of low ethanol consumption. These results indicate that sub- jects’ drinking behavior can be precisely and reliably assessed over a long interval. The finding that reliable self-reports of low ethanol consumption can be obtained from problem drinkers participating in an outcome evalua- tion study is important, since nonproblem drinking is gaining acceptance as an achievable treatment goal for some alcohol abusers. Introduction Although drinking behavior and its possible consequences can now be quantified through objective physiological measurement, the clinical utility of these measures is limited to the verification of recent ethanol consump- tion [l] . To validate reports of past drinking behavior, a convergent validity criterion has been suggested, wherein multiple data sources are evaluated for their mutual corroboration [ 21. However, alcohol abusers’ self-reports remain the most singularly used, easily accessible and convenient measure available to clinicians and researchers for diagnostic and evaluation purposes. A long-standing, but empirically unsupported tenet in the alcohol field is that alcoholics inaccurately report their drinking and drinking-related be- *Reprints should be addressed to Mark B. Sobell at the Clinical Institute, Addiction Research Foundation. @ Elsevier Sequoia/Printed in The Netherlands