Title: Modificações e alternativas aos testes de Levene e Brown e Forsythe para igualdade de variâncias e médias
Abstract: r'he usual tests to compare variantes and means, Bartlett's test and test produced by F-statistic ítom one-way analysis of variante, assume that the sample are ítom normal distributions.Tbhe test for equality of means requires the assumption of homogeneity of variances.There are some problems when these assumptions are not satisfied, like excessivo dize or low power.In this dissertation, we describe the Levene's test for equality of variance, which is robust under nonnormality, and the Brown and Forsythe's test for equality of means.We also present some modifications of the Levene's test and Brown and Forsythe's test, proposed by diHerent authors, and some tests by randomization that can be used as altematives.In addiction, we present an application of the randomized test for equality of means in Bioequivalence problema.Finally, we apply one modiÊied fora of Brown and Forsythe's test to a real data set. Sumário