In Defense of Multiplicative Terms in Multiple Regression Equations

Though the inclusion of multiplicative terms in multiple regression equations is often prescribed as a method for assessing interaction in multivariate relationships, the technique has been criticized for yielding results that are hard to interpret, unreliable (as a result of multicollinearity between the multiplicative term and its constituent variables), and even meaningless. An interpretation of a multiple regression equation with a multiplicative term in conditional terms reveals all these criticisms to be unfounded. In fact, it is better analytic strategy to include a multiplicative term than to exclude one.