Some Graphical Methods for Interpreting Interactions in Logistic and OLS Regression

Abstract

In statistical models involving one dependent variable (DV) and two or more independent variables (IVs), an interaction occurs when the effect of one IV on the DV is different at different levels of another IV. The existence of an interaction makes interpretation of the model more complicated, but failing to include important interactions in the model can give misleading results. In this paper, we describe how visually examining interactions between two IVs in ordinary least squares regression and in logistic regression can aid comprehension of the interaction, and we present a tool to make such examination easier.