Although non-parametric tests have already been proposed for that purpose, statistical significance tests for non-standard measures (different from the classification error) are less often used in the literature. This paper is an attempt at empirically verifying how these tests compare with more classical tests, on various conditions. More precisely, using a very large dataset to estimate the w...

Common using of parametric tests to elaborate research results is limited by predetermined assumptions (variable measurability, normality of its distribution, homogeneity of database etc), which must be fulfilled. Otherwise, the conclusions obtained by calculating using a parametric test will not be quite correct. Parametric tests are useless also in the case of the quality data and the data of...

While performing statistical tests [7] the first question is which statistical test, parametric or non-parametric, is suitable. It is well known that parametric tests have much more power (probability that a statistical test will correctly reject a false null hypothesis, or a probability of avoiding a Type II error) than a non-parametric test. Hence, we should use parametric tests whenever allo...