Essence of non-parametric statistical methods and reasons for which they should be used is explained. Given example shows why significance level should not automatically equal to 5 %.: non-parametric statistics, chi-square, fourfold table, rank statistics, Spearman?s tests of rank correlation