[R-sig-ME] Mixed Models convergence problems, Jordi Rosich
Ben Bolker
bbolker at gmail.com
Sun Dec 4 19:23:00 CET 2016
My other thought about this was that *if* other covariates don't vary
within territory:treatment combinations, and the response is identical
within a territory:treatment combination, you wouldn't lose any actual
inferential power by collapsing to a single value per
territory:treatment combo ... ?
Sorry if that isn't relevant, I haven't been following too closely ...
On 16-12-03 11:58 PM, David Duffy wrote:
>> [In] all our territories, all nests/"no-nests" of a territory are
>> either occupied or unoccupied.
>
> Given your controls are unoccupied trees from within each territory,
> did a conditional logistic regression work (with territory as stratum
> indicator)? You need to either provide a data set (not too big!), or
> speak to a local statistician.
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>
More information about the R-sig-mixed-models
mailing list