Because PQL does not calculate a likelihood, you can't get a deviance.
It's part of the price for a computationally feasible procedure.
On Fri, 16 Aug 2002 msc0235@geo.ed.ac.uk wrote:
> I would like to calculate the percentage explained deviance for models
> obtained with the glmmPQL() procedure, but I do not get any Null deviance or
> residual deviance in the glmmPQL() output. Would anyone know of a way around
> this?
--
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
|