Abstract

Interaction effects refer to the case where the effect of one variable on another depends on the value of a third variable. In this entry, interaction modeling as focused on analysis of variance and multiple regression is introduced using traditional linear models. Interactions are conceptualized using a moderator framework in which the effect of an independent variable or predictor on a dependent or outcome variable is said to vary as a function of a moderator variable. Both two-way and three-way interactions are discussed. Examples are given for the case of all nominal predictors, a mixture of nominal and continuous predictors, and all continuous predictors.