s-news
[Top] [All Lists]

Re: use of stepwise cox regression in comparing relative "importance" of

To: "duo wan" <duo_wan@yahoo.com>, <s-news@lists.biostat.wustl.edu>
Subject: Re: use of stepwise cox regression in comparing relative "importance" of variables
From: "Shawn Boles" <shawn@ori.org>
Date: Wed, 28 Nov 2007 10:21:09 -0800
In-reply-to: <43800.62428.qm@web30302.mail.mud.yahoo.com>
References: <43800.62428.qm@web30302.mail.mud.yahoo.com>
Thread-index: Acgx6A24DWuhMzFhTA+GXHHHeXoI+QAApo0w
Thread-topic: [S] use of stepwise cox regression in comparing relative "importance" of variables

Hi:

 

Frank Harrell identified several problems with using stepwise regression to select variables.

 

See section 4.3 of:

 

Regression Modeling Strategies With Applications to Linear Models, Logistic Regression, and Survival Analysis
Series: Springer Series in Statistics

Harrell, Frank E. Jr.

1st ed. 2001. Corr. 2nd printing, 2006, XXIII, 568 p. 141 illus., Hardcover

ISBN: 978-0-387-95232-1

 

 

 

HTH,

 

Shawn Boles, Ph.D.

Senior Research Associate

Oregon Research Institute

1715 Franklin Blvd.

Eugene, Oregon 97403-1983

USA

Phone (541) 484-2123 ext 2225

Fax: (541) 484-1108

 

 

 


From: s-news-owner@lists.biostat.wustl.edu [mailto:s-news-owner@lists.biostat.wustl.edu] On Behalf Of duo wan
Sent: Wednesday, November 28, 2007 9:56 AM
To: s-news@lists.biostat.wustl.edu
Subject: [S] use of stepwise cox regression in comparing relative "importance" of variables

 

Dear All,

 

I am reading some clinical literatures for our research paper. They adopted stepwise algorithm when building a final prognostic models. However, they used this algorithm for:

 

1)       Comparing the relative importance of two highly correlated variables (different indicators for the same thing). Say the one kept in the model is more important.

2)       Claiming that the variables with higher chi-squares are more important.

 

 

I really do not think two highly correlated variables can be compared with stepwise regression. First, usually one of two highly correlated variables has to go with stepwise regression. It is hard to tell the magnitude of the differences between variables from stepwise.  Also the left one can be more important that some other variables kept in model in terms of higher proportions of explained variations.

 

Is there any good statistical paper about misusing of stepwise in ranking the order of significant of variables or relative importance of variables.

 

Thanks,

 

Vincent

 

 


Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now.

<Prev in Thread] Current Thread [Next in Thread>