Regression Multicollinearity: Sometimes itís helpful

by Robert Lynch and Brian Kim.

Abstract: Examines multicollinearity where collinear variables improve R2 significantly One independent variable is unrelated to the dependent variable and the other has modest to high correlations to the dependent variable and the correlation among the two independent variables is modest to high. An example is provided and a discussion on causes is provided. Focus is on teaching the topic in introductory statistics classes.

Key Words: Linear regression, multicollinearity

Authors:
Robert Lynch, robert.lynch@unco.edu
Brian Kim, jy4860510@gmail.com

Editor: R.G.Graf, rgraf@sunstroke.sdsu.edu

READING THE ARTICLE: You can read the article in portable document (.pdf) format (120550 bytes.)

NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.

This page has been accessed 2438 times since January 29, 2007.


Return to the InterStat Home Page.