scientific article; zbMATH DE number 3974105

From MaRDI portal
Publication:3740071





In this paper we have developed a formula measuring multicollinearity which takes different values for different coefficient estimates and/or for different equations. It takes the correct value of zero for an orthogonal model and a non zero for a nonorthogonal or multicollinear model. Applying this measure of multicollinearity to a few empirical examples, we have found that if one forces a biased regression estimator to satisfy the minimax conditions, then the other goal of reducing multicollinearity may not be realized. We favor choosing from the near minimax estimators one that minimizes the absolute value of the estimate of multicollinearity.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3740071)