How Do You Properly Diagnose Harmful Collinearity In Moderated Regressions?
Source: University of Houston
Collinearity is inevitable in moderated regression models. Most marketing researchers diagnose collinearity using correlation-based metrics such as bivariate correlations and variance inflation factors, and then attempt to decrease the bivariate correlations between the simple effects and the multiplicative interactions to alleviate the collinearity problems. This paper clearly demonstrates through both graphical proofs and a numerical example that it is possible to have highly collinear relationships in a model, yet have negligible bivariate correlations among the individual variables. Paper illustrates the usefulness of the C2 metric using an example from the brand extension literature. Finally, paper provides best practice recommendations for researchers to diagnose and manage collinearity related issues.