single argument anova for GLMMs not yet implemented
On Thu, 2008-12-11 at 14:58 -0600, Douglas Bates wrote:
On Thu, Dec 11, 2008 at 2:52 PM, Andrew Robinson <A.Robinson at ms.unimelb.edu.au> wrote:
Echoing Murray's points here - nicely put, Murray - it seems to me that the quasi-likelihood and the GLMM are different approaches to the same problem.
I agree and I also appreciate Murray's elegant explanation.
Can anyone provide a substantial example where random effects and quasilikelihood have both been necessary?
I'm kind of waiting for Ben Bolker to let us know how things look from his perspective. I seem to remember that Ben and others in ecological fields were concerned about overdispersion, even after incorporating random effects.
Not wanting to preempt Ben or anything, but yes, we ecologists are very concerned about overdispersion. However, and I say this as someone new to this field (mixed models, not ecology), the quasilikelihood approach seems far more of a fudge to avoid having to think about where the overdispersion is coming from. I find the negative binomial far more intuitive to deal with than working around the problems not having a proper likelihood brings (inference, model selection, information stats). Often, the course of overdispersion is of direct interest itself. In the GLM arena I find hurdle and ZIP and ZINB models far more interpretable in ecological terms than fudging the problem with quasilikelihood methods. And after-all, that is what I am interested in; models I can interpret in ecological terms. My two-penneth, G
Best wishes, Andrew On Fri, Dec 12, 2008 at 09:11:39AM +1300, Murray Jorgensen wrote:
The following is how I think about this at the moment: The quasi-likelihood approach is an attempt at a model-free approach to the problem of overdispersion in non-Gaussian regression situations where standard distributional assumptions fail to provide the observed mean-variance relationship. The glmm approach, on the other hand, does not abandon models and likelihood but seeks to account for the observed mean-variance relationship by adding unobserved latent variables (random effects) to the model. Seeking to combine the two approaches by using both quasilikelihood *and* random effects would seem to be asking for trouble as being able to use two tools on one problem would give a lot of flexibility to the parameter estimation; probably leading to a very flat quasilikelihood surface and ill-determined optima. But all of the above is only thoughts without the benefit of either serious attempts at fitting real data or doing serious theory so I will defer to anyone who has done either! Philosophically, at least, there seems to be clash between the two approaches and I doubt that attempts to combine them will be successful. Murray Jorgensen
-- Andrew Robinson Department of Mathematics and Statistics Tel: +61-3-8344-6410 University of Melbourne, VIC 3010 Australia Fax: +61-3-8344-4599 http://www.ms.unimelb.edu.au/~andrewpr http://blogs.mbs.edu/fishing-in-the-bay/
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% Dr. Gavin Simpson [t] +44 (0)20 7679 0522 ECRC, UCL Geography, [f] +44 (0)20 7679 0565 Pearson Building, [e] gavin.simpsonATNOSPAMucl.ac.uk Gower Street, London [w] http://www.ucl.ac.uk/~ucfagls/ UK. WC1E 6BT. [w] http://www.freshwaters.org.uk %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%