Abstract
Synopsis A sample of male prisoners was examined by means of a questionnaire and subsequently by a structured interview. Control subjects who denied heavy drinking were used to study test-retest reliability of a group who admitted that drinking was a problem and a third group who admitted heavy drinking but denied problem drinking. Non-heavy drinkers and the heaviest drinkers are shown to be the most reliable and the intermediate group least reliable.

This publication has 48 references indexed in Scilit: