[R-sig-ME] Model validation for Presence / Absence, (binomial) GLMs
Ben Bolker
bbolker at gmail.com
Thu Jul 4 19:55:46 CEST 2013
Highland Statistics Ltd <highstat at ...> writes:
>
>
> >> This is something I always battle with given the plethora of great model
> >> fitting methods available for other models.
> >>
> >> I always use a variant of Hugh's suggestion and look at the % of correct
> >> predictions between models as a quick model fitting statistic.
> >>
> >> And for overdispersion I believe one way is to fit individual level random
> >> effects and see if this is a substantively better model. There is more on
> >> this in the wiki http://glmm.wikidot.com/faq
> > Yes, but this is unidentifiable for Bernoulli responses (as also
> > explained there).
>
> The statement on 'unidentifiable for Bernoulli
> responses'....well...apparently this is not that trivial.
> See: http://www.highstat.com/BGGLM.htm
>
> Follow the link to: the Discussion Board....
> Go to: Chapter 1 Introduction to generalized linear models
> And see the topic: Can binary logistic models be overdispersed?
>
> Alain
That's an interesting document: I think the bottom line is:
* if the Bernoulli data can be grouped, i.e. if there are
in general multiple observations with the same set of covariates,
then overdispersion can be identified, because the data are
really equivalent to a binomial response within the groups.
For example, the trivial example
grp resp
A 1
A 0
A 1
B 0
B 0
B 1
is equivalent to:
grp successes total
A 2 3
B 1 3
this sort of aggregation can be done with plyr::ddply, or
in lots of different ways -- it's generally faster and
more efficient to fit the data in the grouped form than
in the binary form.
It might have been more convenient for readers to be able to
go directly to the link:
http://www.highstat.com/BGS/GLMGLMM/pdfs/
HILBE-Can_binary_logistic_models_be_overdispersed2Jul2013.pdf
{URL broken to make Gmane happy}
rather than having to click 5 times to get through the
Highland Statistics material to the PDF of interest ...
More information about the R-sig-mixed-models
mailing list