s-news
[Top] [All Lists]

Re: proportional odds assumption

To: "Coppenolle, Hans" <Hans.Coppenolle@AGR.KULEUVEN.AC.BE>
Subject: Re: proportional odds assumption
From: Prof Brian Ripley <ripley@stats.ox.ac.uk>
Date: Tue, 31 Jul 2001 17:48:40 +0100 (BST)
Cc: internet <s-news@lists.biostat.wustl.edu>
In-reply-to: <D8BDEB41755B704690C96ADD78B0E1F412C856@agr-srv-24.agr.ad10.intern.kuleuven.ac.be>
On Tue, 31 Jul 2001, Coppenolle, Hans wrote:

> Does the possibility exist in s-plus 2000
> to test the proportional odds assumption
> when fitting a proportional odds model?
>
> The function polr seems not to include that
> property...

Take a look in MASS (the book) for a worked example ....

> How can the goodness of fit of a logistic proportional odds
> model and a generalized (baseline) logistic model
> efficiently be compared??

Um: logistic proportional odds are for K > 2 classes, and compare with
multiple logistic models.

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595


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