s-news
[Top] [All Lists]

Re: no p-values

To: Bernd Puschner <bernd.puschner@bkh-guenzburg.de>
Subject: Re: no p-values
From: Prof Brian Ripley <ripley@stats.ox.ac.uk>
Date: Fri, 18 Jul 2003 11:27:21 +0100 (BST)
Cc: snews <s-news@lists.biostat.wustl.edu>
In-reply-to: <JHEILOKJLDODCBAOAKDOOEGCCEAA.bernd.puschner@bkh-guenzburg.de>
On Wed, 16 Jul 2003, Bernd Puschner wrote:

> when computing a logistic regression model, either via glm or
> menuBinomialGlm, the summary of the model provides coefficients and their
> standard errors and t-values, but for some reason no significance tests
> (p-values) which i would like to have. any ideas?

There is a very good reason!  t-values for logistic regression can be
highly misleading: small values indicate either a small effect or a very
large effect (the `Hauck-Donner' phenomenon).  The distribution theory is 
far too complicated, as the asymptotic approximations are not good enough.
(the t-values are not t-distributed).

Use dropterm() in library MASS to obtain likelihood-ratio tests for each 
term in the model: those are more reliable.  Also use profiling (?profile) 
to obtain confidence intervals.

-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595


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