Resumen

The Wilcoxon-Mann-Whitney (WMW) test is frequently used to compare means and medians of two independent sets with a possible non-normal distribution. This is not correct and can lead to an incorrect analysis. The difference in the data spread of one group compared to another is established by the WMW test. We emphasize the adequate use of this test. The WMW test is used as an alternative of two independent samples of the t-Student test. An analysis was carried out of 100 indexed and non-indexed, medical articles, national, and international, and it was observed that the WMW test was used unequivocally. Thus, the WMW test is not exactly a medians test, but what is important is the spread difference. The WMW test allows us to reach different conclusions that depend on the shape and distribution of data. The non-parametric methods are based on the ordering of data by magnitude and to replace these data by rank counting.