Efficacy of Centering Techniques for Creating Interaction Terms in Multiple Regression for Modeling Brand Extension Evaluation

Abstract

Interaction terms (representing interaction of the model variables) are commonplace when it comes to multiple regression. Our primary contribution is to verify the benefit of applying different statistical techniques for creating interaction effects in multiple regression models, specifically for the case of brand extension model. Cross product interaction term is considered as a traditional technique for creating interaction effects as compared to mean centered interaction term or the residual centered interaction term. Prior research indicated the problem of multicolinearity in using cross product interaction term and suggested use of the recent analytical techniques, which involved centering of the variables. In the current study, it was found that centering only helps multicollinearity disappear and doesn’t quite improve the regression model as such.