[Experimental statistical studies on the evaluation of the value of evidence in the comparison of human hair].

In this paper a statistical experiment is presented which allows to estimate the evidential value of human head hair comparison. The procedure described in detail was essentially the same as in case-work. Questioned hairs were taken randomly from 20 different persons out of a pool of 111 individuals. From each of the 20 persons one, three an five hairs respectively were compared consectively with samples from 100 different individuals. The results were classified as follows: a) "matching" and b) "similar" - if the hair(s) may originate form that person; c) "not matching" - if an individual is excluded as possible source. In our experiment about 95% of the samples (persons) were excluded as possible source of a questioned hair on the average. This is a mean value which may vary considerably in a distinct case. The experiment, its results and problems are discussed.