Control Variables in Interactive Models

Accounting studies often examine whether the relation between X and Y varies with a moderating variable, M, by including an interactive term, X×M, in a regression. We provide plain-English guidance on why, how, and when to use control variables (Z) in interaction tests. Simple descriptions and simulations demonstrate the roles of control variables and how interacted controls affect coefficient estimates and interpretations. In particular, we demonstrate how including Z without an accompanying interaction of X×Z and/or M×Z generally does not eliminate the confounding effect of Z on X×M. We conclude with guidance for future research