Date:         Mon, 19 Jul 2010 11:31:26 +0000
Reply-To:     goladin@gmail.com
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "goladin@gmail.com" <goladin@GMAIL.COM>
Subject:      Re: stepwise and colinearity
Comments: To: Leonardo Auslender <Leonardo.Auslender@sas.com>
Content-Type: text/plain; charset="utf-8"

Hi,

Agreed. The problem is that people believe that stepwise selection removed collinearity when it might cause it.

Regards,
Murphy Choy
Sent from my Nokia phone
-----Original Message-----
From: Leonardo Auslender
Sent:  19/07/2010 7:30:02 PM
Subject:  stepwise and colinearity

Hello,

I can't answer directly to SAS-L. Stepwise does reduce quite a bit of colinearity but the resulting model could still be collinear because the inference path followed by stepwise leads to redundancy.