Mathematical Researches

The homogeneity of variances test is a prerequisite for many statistical methods. In this article, a recently introduced test based on the jackknife approach is compared with common tests such as Levene's and Bartlett's tests, as well as two tests by James and Alexander-Govern, in terms of their ability to maintain the first type error rate and test power for several distributions. The permutation versions of these tests were also examined. The results indicate that the performance of the tests significantly improves in the permutation version. To evaluate the performance of the tests in the real world, the tests were applied to two real data sets, and the results are presented.