Regression Estimation Accuracy: Multicollinearity, Explanation, and Prediction

• In Event: Missing Data, Multicollinearity, and Variable Selection in Regression

Abstract

Recommendations from popular statistics texts regarding avoidance of predictor variable multicollinearity in the use of multiple regression are considered from the perspective of the alternate purposes of explanation and prediction. For 21 data sets examined, the different effects of multicollinearity on prediction and explanation accuracy are presented. Prediction accuracy (relative or absolute) was not affected by multicollinearity; but multicollinearity manifested an appreciable effect on explanation accuracy. Consideration of the degree of effect at differing levels multicollinearity in respect to orthogonality and an alternative (ridge regression) that may partially ameliorate this loss in explanation accuracy are also considered.

Authors

• Mary G. Lieberman, Florida Atlantic University

• John D. Morris, Florida Atlantic University