The effect of the normality and homogeneity of variance assumptions upon the validity of the F table for interaction terms

The purpose of this thesis is to investigate the validity of the P Table under the following conditions,first J when the assumption of common variance is violated and, secondly5 when both the normality and homogeneity of variance assumptions are violated. Using a factorial design, with special interest in the effects upon interactions, this investigation is concerned with the validity of the P Table by simulating experiments with the data source being from different distributions, each containing an uncommon variance.