The Assumptions Underlying the Analysis of Variance

Abstract
An explicit statement of the assumptions underlying the analysis of variance and the practical importance of each. Models of the 2 distinct classes of problems solvable by analysis of variance are discussed: a) detection and estimation of fixed relations among the means of subsets of the universe of objects concerned, and b) detection and estimation of components of variation associated with a composite population. The reasons for the assumptions made for each class are explained, as well as the choice of the model appropriate for a given problem.
Keywords

This publication has 2 references indexed in Scilit: