Two-way analysis of variance using the regression model

Abstract

The analysis of variance model can be extended to situations involving two or more categorical independent variables. These more complex analyses of variance can be performed using regression models with dummy variables. In order to understand the logic of these models, we can examine the case of two categorical independent variables. An analysis of variance with two independent variables is referred to as a two-way analysis of variance. Of course, this model can easily be extended to handle additional categorical independent variables. In any event, the approach is similar to that employed in the simple one-way analysis of variance. In a two-way analysis of variance using regression analysis with dummy variables, binary variables are created to represent all but one of the categories of each categorical independent variable.