Generalized Collinearity Diagnostics Revisited

• In Event: Multiple Linear Regression: The General Linear Model SIG Paper Session: Statistical Modeling and Related Issues in the Implementation of Empirical Educational Data

Abstract

Fox and Monette (1992) introduced the Generalized Variance Inflation Factor (GVIF) as a means of assessing multicollinearity for terms in linear models that have more than 1 degree-of-freedom. This applies to models containing related sets of regressors including: (1) a set of coding scheme variables that represent independent groups (i.e., one-way; factorial between-subjects ANOVA designs) or (2) variables that are functions of each other as in polynomial regression and response surface analysis. Despite this important development, GVIF rarely appears in the statistical literature or statistical software. The purpose of this paper is to: review the issues of multicollinearity; reintroduce GVIF; and demonstrate how to compute these values in popular statistical software (i.e., SAS, SPSS, STATA) that do not report GVIFs.

Authors

• T. Mark Beasley, University of Alabama - Birmingham
  Presenting Author

• Amandiy N Liwo, University of Alabama - Birmingham
  Non-Presenting Author