| pubmed-article:3969843 | pubmed:abstractText | The concurrent validity of questionnaires in population studies on drug use was tested by comparing information from registers and interviews in two populations of young people from Gothenburg. The questionnaire items' sensitivity and specificity according to criteria and their capacity for prediction of drug use and no drug use and correct classification were calculated. Considering criteria weaknesses and other sources of errors, the validity of the questionnaire items was good. This conclusion is basic to longitudinal population studies which aim at clarifying the prognosis of drug abuse among young people. | lld:pubmed |
