[Use of nonparametric methods in medicine. VIII. Conclusion].

The submitted paper terminates the series of eight articles on the application of non-parametric methods in medicine, which were published in this journal in 1990-1991. Attention is drawn, on the one hand, to the urgent need of knowledge of mathematical and statistical methods, on the other hand, to the fact that proper processing of experimental data is not guaranteed by the use of computer technique alone, the complexity and numerical correctness of calculations but by interpretation of results and their plausibility which depends above all on the correctness of the reflections of the research worker. On a general level attention is paid to the description, comparison and evaluation of correlations, in particular the possibility to conclude from statistical correlation that a causal relationship exists. In the present as well as in previous articles the authors emphasize repeatedly that statistics are only one of the important tools of the doctor, but they cannot replace his responsibility, nor his professional experience.