Date: Tue, 17 Apr 2001 14:01:38 -0400
Reply-To: Peter Flom <peter.flom@NDRI.ORG>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Peter Flom <peter.flom@NDRI.ORG>
Subject: Re: Multicollinearity in Logistic Regression
Multicollinearity can affect any regression procedure. One way to look at it in logistic in SAS is to "fool" the computer into thinking you are doing regular regression, and use the /collin option. Since collinearity is a relationship among the independent variables, it's irrelevant that PROC REG is inappropriate for your dependent variable
Peter L. Flom, Ph.D.
Principal Research Associate
National Development and Research Institutes, Inc.
2 World Trade Center
New York, NY 10048
(212) 845-4698 (fax)
>>> "Wang, Zhaoping (PCM - NJ)" <ZWang@NA2.US.ML.COM> 04/17/01 01:28PM >>>
Is there the issue of multicollinearity that we should
look at when running a logistic regression ? (I know
there is in a linear regression.) If yes, how to
take care of it in SAS ? Thanks for your help.