ABSTRACT
The basic purpose of analysis of variance (ANOVA) is to assist the researcher in formulating a linear model that is appropriate for describing the data obtained in a study. The most appropriate model is one that is as simple as possible, yet still provides an adequate description of the data. Although the simplicity and adequacy of a particular model could be evaluated on an absolute basis, typically models are judged on a relative basis by comparisons with other possible models. This notion of searching for a simple yet adequate model is pervasive. It informs not only all applications of ANOV A, but also many other kinds of hypothesis testing.
