LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (July 2008, week 3)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Fri, 18 Jul 2008 16:18:09 -0500
Reply-To:     Mary <mlhoward@avalon.net>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Mary <mlhoward@AVALON.NET>
Subject:      Re: diagnosis of multicollinearity in event history analysis
Comments: To: Anna Cui <cuis@RPI.EDU>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
              reply-type=original

Anna,

One approach might be to use cluster analysis procedures like VARCLUS just on your independent variables to see if there are groups of them that cluster (and therefore are likely targets for multicolliarity); if they do then you might consider using the cluster score for that set of correlated variables rather than the original variables. Here is an example of using VARCLUS:

ods output RSquare=RSquare_set; proc varclus data=set2 outstat=Coef; var var1--var100; run;

ods select all;

data Coef2; set Coef; if _ncl_ = 10; drop _ncl_; run;

proc score data=set2 score=Coef2 out=scoreset;

run;

-Mary

----- Original Message ----- From: Anna Cui To: SAS-L@LISTSERV.UGA.EDU Sent: Friday, July 18, 2008 2:53 PM Subject: diagnosis of multicollinearity in event history analysis

I know when i run regression with proc reg, I can ask for collinearity diagnosis statistics such as VIF and condition index, to see whether I have multicollinearity problems. But when I am using proc phreg, how can I get these? Or there are different ways to test for multicollinearity?

Thank you very much.

Anna


Back to: Top of message | Previous page | Main SAS-L page