A psychometric evaluation of an instrument designed to measure alcohol-induced blackouts

Sidney Guy Fooshee

Abstract

The Alcohol Blackout Questionnaire is currently the only instrument available for alcohol blackouts research. A psychometric evaluation was conducted on the instrument and suggestions for its revision and improvement are presented. The questionnaire has a six factor structure which accounts for 59.0% of the variance. The first factor was a general blackout factor. The second factor related to thoughts and feelings. The third factor involved high risk behaviors. The fourth factor related to drug use. The fifth factor related to the person who realized they had a problem while the sixth factor addressed people who did not realize they had a problem with alcohol. A step-wise discriminant analysis revealed 19 questions which were the best predictors of alcohol blackouts. Overall the instrument correctly classified 88.28% of the individuals. The blackout group had a classification rate of 91.7% while the nonblackout group had a classification rate of 86.3%. Suggestions are presented concerning the restructuring of the questionnaire. Instructions are presented at the top of every page. The clinical history type questions now appear at the front of the questionnaire and the blackout questions at the end. The rating scales are more concise and easily understood.