Bootstrap graphical test for equality of variances

Abstract

Testing hypothesis about variance parameters arises in contexts where uniformity is important and also in relation to checking assumptions as a preliminary to analysis of variance, discriminant analysis and so forth. In this paper, a bootstrap graphical method is developed as an alternative to the Bartlett and Levene tests to test the hypothesis on equality of several variances. An example is given to demonstrate the advantage of bootstrap graphical procedure over the Bartlett and Levene tests from decision making point of view.