Validity of children's reports of parental alcohol abuse

Data from interviews with parents and 9-12 year old children from 303 families were used to test the validity of children's reports of parental alcoholism or problem drinking status. Children's responses to several single-item screening questions that had been used previously to identify children of alcoholics, as well as their responses to the Short Michigan Alcoholism Screening Test for each parent, were compared to parents' alcoholism diagnosis and screening test scores. In general, children's responses correctly identified only a small portion of alcoholic parents and, when all screening items were combined into a single screening tool to increase sensitivity, the false positive rate became unreasonable for most applications. The screening items were better at identifying active alcoholics than abstaining alcoholics but were also better at identifying problem drinking parents than alcoholic parents. None of the screening options tested had optimal characteristics for use in research or intervention applications.