Article View: sci.math.stat
Article #766Re: stepwise regression
From: Frank Harrell
Date: Fri, 18 Dec 1992 02:15
Date: Fri, 18 Dec 1992 02:15
12 lines
964 bytes
964 bytes
I very much agree with Scott Thompson's thoughful post. More often than not, the best prediction arises from a full model fit (i.e. collinearities aren't really all that dangerous) followed by shrinkage. Spiegelhalter (Stat in Med 5:421-33, 1986) says this better than I could. -- ---------------------------------------------------------------------------- Frank E Harrell Jr feh@biostat.mc.duke.edu Associate Professor of Biostatistics Division of Biometry Duke University Medical Center
Technical Details
Message-ID:
<8120@news.duke.edu>Path:
arc1-ca.pugleaf.net!sparky!uunet!olivea!spool.mu.edu!agate!stanford.edu!rock!concert!duke!news.duke.edu!duke.edu!fehReferences:
<8091@news.duke.edu> <thompson.724621440@daphne.socsci.umn.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!olivea!spool.mu.edu!agate!stanford.edu!rock!concert!duke!news.duke.edu!duke.edu!feh From: feh@duke.edu (Frank Harrell) Newsgroups: sci.math.stat Subject: Re: stepwise regression Summary: amen Message-ID: <8120@news.duke.edu> Date: 18 Dec 92 02:15:30 GMT References: <8091@news.duke.edu> <thompson.724621440@daphne.socsci.umn.edu> Sender: news@news.duke.edu Lines: 12 Nntp-Posting-Host: biostat.mc.duke.edu