SummaryThe therapeutic orientation tests, which are indispensable during the preliminary stage, need confirmation through severely controlled, more individual tests. In case where strictly objective evaluation criteria are lacking, a comparison between the new drug and a placebo or a standard-drug seems necessary; the formation of comparable groups without arbitrary selection and the choice of a depersonified psychological approach of the patient will allow the clinician to estahlish a correct statistical analysis of his results. The experimental scheme and the adequate statistical method will be selected with regard to the practical organization possibilities.