Alcohol abusers' perceptions of the accuracy of their self-reports of drinking: implications for treatment.

Several major literature reviews have concluded that alcohol abusers generally give valid self-reports when interviewed under certain conditions. Nevertheless, across all studies a small proportion of alcohol abusers' self-reports continue to be suspect. Sources of invalidity may relate to subject factors or to circumstances under which data are collected. One novel way of gaining information about conditions possibly affecting the accuracy of alcohol abusers' self-reports is to ask the subjects themselves. In the present study, 208 alcohol abusers were asked about (a) how accurately different people they knew or lived with would report their (i.e., the subjects') drinking at different levels (e.g., abstinent, 1-4 drinks) compared to the subjects' own reports; (b) how accurate their own reports would be at different levels of ethanol consumption; and (c) how accurate their own reports would be when interviewed under different conditions (e.g., by phone, their therapist, a researcher, their employer). The results are largely consistent with studies that have empirically examined the validity of alcohol abusers' self-reports. Suggestions for future research and evaluation are offered.