Analysis of Variance (ANOVA) is an important technique to test the homogeneity of means for several populations. One of the basic assumptions for ANOVA is that the variances of several populations are equal. In the real world, this kind of assumption may or may not be true. One needs to check the validity of this assumption before applying the technique of ANOVA. Several tests have been proposed for the homogeneity of variances by Conover et al. [4] in Completely Randomized Design (CRD) and so far, the most frequently cited and used are methods proposed by Bartlett [1], Box [3], Hartley [9], Levene [14]. Bhandary and Dai [2] recently proposed an alternative test in this situation. Harris [8] developed a variance homogeneity test for correlated variables, Piepho [16–19] proposed several methods to test homoscedasticity.
Analysis of Variance (ANOVA) is an important technique to test the homogeneity of means for several populations. One of the basic assumptions for ANOVA is that the variances of several populations are equal. In the real world, this kind of assumption may or may not be true. One needs to check the validity of this assumption before applying the technique of ANOVA. Several tests have been proposed for the homogeneity of variances by Conover et al. [4] in Completely Randomized Design (CRD) and so far, the most frequently cited and used are methods proposed by Bartlett [1], Box [3], Hartley [9], Levene [14]. Bhandary and Dai [2] recently proposed an alternative test in this situation. Harris [8] developed a variance homogeneity test for correlated variables, Piepho [16–19] proposed several methods to test homoscedasticity.
การแปล กรุณารอสักครู่..
![](//thimg.ilovetranslation.com/pic/loading_3.gif?v=b9814dd30c1d7c59_8619)