Date:         Wed, 17 Jul 1996 12:44:41 -0600
Reply-To:     "Raymond V. Liedka" <liedka@UNM.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         "Raymond V. Liedka" <liedka@UNM.EDU>
Subject:      Re: GENMOD procedure
Comments: To: Emmanuel Jeudy <jeudye@ULB.AC.BE>
In-Reply-To:  <m0ugRLy-00021TC@alcor.unm.edu>

On Mon, 15 Jul 1996, Emmanuel Jeudy wrote:
 
> I am trying to deal with this procedure because I am actually dealing
> with a subject very close to regression with ordinal response variable
> (McCullag, 1980). But unfortunatly I had not been yet able to find the
> (paper) documentation, which is much more convenient than the on-line
> one.
 
If you are actually looking at the models in the 1980 McCullagh paper on
regression models for ordinal variables, you will actually be better off
using PROC LOGISTIC.  That procedure fits the cumulative logit (aka the
proportional odds) model that gets a great deal of attention in the
paper.  This proc is detailed in the 2-volume Statistics manuals.
 
 
Raymond V. Liedka
Department of Sociology
University of New Mexico