[19185] in s-news-athena
Re: [S] binary responses
daemon@ATHENA.MIT.EDU (Prof Brian D Ripley)
Sat Sep 11 03:35:02 1999
Date: Sat, 11 Sep 1999 08:24:15 +0100 (BST)
From: Prof Brian D Ripley <ripley@stats.ox.ac.uk>
To: James Lymp <lymp@u.washington.edu>
Cc: "Frank R. Lawrence" <Cougar@UAB.EDU>, s-news@wubios.wustl.edu
In-Reply-To: <008001befbc5$9689e0a0$cecf8e8c@u.washington.edu>
Message-Id: <Pine.GSO.4.05.9909110806120.9715-100000@auk.stats>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
On Fri, 10 Sep 1999, James Lymp wrote:
>
> "Generalized Linear Mixed Models (GLMM)" and "Generalized Estimating
> Equations (GEE)" are two methods for modeling repeating measures, depending
> on the nature of the question. Both methods extend the "Generalized Linear
> Model (GLM)". I am not sure how to fit GLMMs in S-PLUS, however, there is a
> gee package on statlib (http://lib.stat.cmu.edu/S/). I don't believe that
> gee has been officially incorporated into S-PLUS yet.
The question was about S-PLUS 2000. There are two versions on
statlib for Windows, and only one (.../DOS/S/SWin) works on 2000.
However, I don't think gee is really about repeated measures at all. It is
about correlated outcomes, in this case Bernoulli, and indeed with
specified correlation structure. Correlations are not the most natural way
to handle dependence in binary variables. Repeated measures designs are to
my mind most naturally handled by random effects, that is GLMMs here.
With other distributions (normal, binomial, Poisson) the natural models
give marginal distributions that are not (normal, binomial, Poisson), and
it is the marginal distributions that gee models. This is not a problem
with Bernoulli, though, as mixtures of Bernoulli are Bernoulli.
There are simpler approaches to repeated measures. If the interest is in
the two treatments, I would start by fitting a logistic regression
with a per-subject factor, and look for differences in the response
patterns by group. Something like
glm(success ~ factor(subject) + time*treatment, family=binomial, ...)
assumming treatment and time (the three times of response) are factors. and
see if the treatment x time and treatment terms are (practically)
significant.
This is just like in the normal case using a fixed-effect randomized block
analysis for treatment differences, even though the model is of random
effects for blocks.
[I would need a lot more details on the experiment before actually doing
things like this, though, and suggest you talk to the local statistical
consultant.]
>
> -----
> James Lymp
> http://staff.washington.edu/lymp
> lymp@u.washington.edu
> -----
>
> ----- Original Message -----
> From: Frank R. Lawrence <Cougar@UAB.EDU>
> To: <s-news@wubios.wustl.edu>
> Sent: Thursday, September 09, 1999 2:45 AM
> Subject: [S] binary responses
>
>
> I would like to know how to model dichotomous repeated measures using S-Plus
> 2000. The situation is there are three dichotomous responses collected from
> 250 subjects. The 250 subjects were randomly assigned to two treatment
> conditions.
>
>
> -----------------------------------------------------------------------
> This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
> send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
> message: unsubscribe s-news
>
--
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
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
