Date:         Tue, 6 Jun 2006 12:12:00 -0400
Reply-To:     Kevin Roland Viel <kviel@EMORY.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Kevin Roland Viel <kviel@EMORY.EDU>
Subject:      Re: Question/rant about AIC and PROC REG and so on
In-Reply-To:  <BAY103-F35D3B4A2EF5290D18F71A4B0950@phx.gbl>
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Mon, 5 Jun 2006, David L Cassell wrote:

> The SELECTION= bit is annoying.  But OTOH, why would people
> be after the AIC or Mallows' Cp unless they were fitting more than
> one model?  Under normal circumstances, AIC seems like non-useful
> clutter when fitting a single model in PROC REG.  If you have K
> variables and you insist on START=k and STOP=k, then you only
> get one model fit.  But I think that's a kludge.

I have used the AIC to compare different models that were not simply
subsets or supersets of the others.  In that case, I was interested in
different representations of the same factor-broadly discrete versus
continous.

The AIC is not difficult to calculate by hand, one just needs the
likelihood and the (total) number of parameters.  It is important to
stress, though, that the AIC pertains on to the collection of data at
hand.  One cannot compare AIC between different studies...

Kevin

Kevin Viel
Department of Epidemiology
Rollins School of Public Health
Emory University
Atlanta, GA 30322