Validity of alcoholics' self-reports: Duration data

The validity of alcohol abusers' self-reports of the number of days associated with alcohol-related incarcerations and hospitalizations was investigated using a recently developed time-line follow-back interviewing technique. It was felt that duration data (number of days per occurrence) would be more sensitve than incidence data (number of occurrences) to certain types of invalidity. Results indicated that the majority of both residential and outpatient subjects provided relatively accurate self-reports of the number of days they had been incarcerated during the 360 days preceding their entry into treatment. Discrepancies most frequently resulted from subjects overreporting days incarcerated as compared with official records.