Robustness of power in the analysis of variance

Abstract

The object of the present work is to study the sensitivity of the power in Anaiysis of variance in relation to the departure from the usual test assumptions (i) equality of variance of the errors, (ii) statistical independence of the errors, and (iii) normality of the errors. In order to obtain the power value in relation to the departure from the usual test assumptions other than the normality of errors, the most General Linear Hypothesis model is considered. The power values so obtained are presented in Tables I and II. The result suggests that in the above model for the inference about means, the power value is greatly affected by the inequality of error variances but only slightly affected by the serially correlated error variables