[R] Mixed effects model with binomial errors - problem

RFTW l.temarvelde at nioo.knaw.nl
Wed Sep 10 16:13:49 CEST 2008


ok, the model does run now! 

but, dont i need the method="ML" when i want to compare this model with a
reduced model  using anova(model1, model2)? The R-Book tells me that REML is
not good for that (p.635)

So, besides that... how do i now do a sort of posthoc test to see 1)
estimates of all treatments and 2) which treatments are different from
which.

Luc. 
PS. the names are not the original names in the file






Ben Bolker wrote:
> 
> RFTW <l.temarvelde <at> nioo.knaw.nl> writes:
> 
> 
>> Our question is if the proportion of visits to the experimental tree, in
>> relation to the total number of visits to both trees differs between
>> treatments.
>> 
>> We have made treatment and individual into a factor
>> All individuals were subjected to a maximum of 4  treatments, so
>> 'individual' is a random factor
>> 
>> We came up with this model:
>> model1<-lmer(cbind(ExpTree,Total visits-ExpTree)~ Treat
>> +(1|Ind),method="ML"
>> , family=binomial, data=r))
> 
>   why not leave out method="ML" and see what happens?
>   for the current iteration of lmer, REML is not a possibility
> in any case.  The default "Laplace" method should work OK.
> 
>    I'd be slightly worried about your variable name with a space
> in it (`Total visits`) -- are you sure that is working as expected?
> 
>   For further questions along these lines I would suggest
> e-mailing r-sig-mixed-models at lists.r-project.org instead ...
> 
>   Ben Bolker
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
> 

-- 
View this message in context: http://www.nabble.com/Mixed-effects-model-with-binomial-errors----problem-tp19413327p19414516.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list