[R-sig-ME] Confidence intervals for effects in glmer()

[Ben Bolker]

  For a while mcmcsamp was withdrawn/labeled as untrustworthy by Doug
Bates, I'm not sure what its current status is.  In any case, as far as
I know it has never worked for GLMM (as opposed to LMM) fits -- which
is what the original poster asked about.  If it does work, mcmcsamp
is definitely better than the solutions I proposed.

  Ben Bolker


CL Pressland wrote:
> Dear fellow novice,
> 
> I've used mcmc to estimate confidence intervals in lme4. You don't need the 
> coda package as it's included in lme4 now. For you it would look like this:
> 
> fit4.mcmc<-mcmcsamp(fit4, n=1000) #or whatever resample number you think is 
> appropriate
> HPDinterval(fit4.mcmc)	#this will fit to alpha 0.05 as default.
> 
> As far as I am aware this is a fairly good method of estimating CIs in 
> mixed models, but Ben are you saying that there may be issues with this 
> process?
> 
> Kate
> 
> --On 08 June 2009 19:40 -0400 Ben Bolker <bolker at ufl.edu> wrote:
> 
>>   It depends on what approximations you're willing to accept.
>>
>>   Crudely doing +/- 2 standard deviations, or (DANGER DANGER)
>> using the 'known' df to compute t-scores (the "gm1" example
>> given in ?glmer has a fairly straightforward structure, with
>> 15 groups):
>>
>> library(lme4)
>> example(glmer)
>> s <- summary(gm1)@coefs
>>
>> fac <- 2
>> s[,"Estimate"]+fac*outer(c(-1,1),s[,"Std. Error"])
>>
>> fac <- qt(0.975,df=14)
>> s[,"Estimate"]+fac*outer(c(-1,1),s[,"Std. Error"])
>>
>>
>>   Does anyone know the current status of mcmcsamp,
>> either for LMMs or for GLMMs ... ?
>>
>>   Ben Bolker
>>
>> Robert A. LaBudde wrote:
>>> This may be blindingly obvious to the casual observer, but I'm
>>> chagrined to admit I'm stumped.
>>>
>>> I'm fitting a simple mixed effect logistic model using 'lme4':
>>>
>>> require('lme4')
>>> fit4<- glmer(x ~ 1 + 1|lab, data=eg, nAGQ=5, family='binomial')
>>> summary(fit4)
>>> ranef(fit4)
>>>
>>> I would like 95% confidence intervals on 'lab' and the residuals effects.
>>>
>>> Using lme() in 'nlme', I had the function intervals() available. Now I
>>> don't.
>>>
>>> Any hints to de-perplex a novice?
>>>
>>> Thanks.
>>> ================================================================
>>> Robert A. LaBudde, PhD, PAS, Dpl. ACAFS  e-mail: ral at lcfltd.com
>>> Least Cost Formulations, Ltd.            URL: http://lcfltd.com/
>>> 824 Timberlake Drive                     Tel: 757-467-0954
>>> Virginia Beach, VA 23464-3239            Fax: 757-467-2947
>>>
>>> "Vere scire est per causas scire"
>>>
>>> _______________________________________________
>>> R-sig-mixed-models at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>>
>> --
>> Ben Bolker
>> Associate professor, Biology Dep't, Univ. of Florida
>> bolker at ufl.edu / www.zoology.ufl.edu/bolker
>> GPG key: www.zoology.ufl.edu/bolker/benbolker-publickey.asc
>>
>> _______________________________________________
>> R-sig-mixed-models at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
> 
> 
> 
> ----------------------
> Kate Pressland
> Office D95
> School of Biological Sciences
> University of Bristol
> Woodland Road
> Bristol, BS8 1UG
> 
> Kate.Pressland at bristol.ac.uk
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models


-- 
Ben Bolker
Associate professor, Biology Dep't, Univ. of Florida
bolker at ufl.edu / www.zoology.ufl.edu/bolker
GPG key: www.zoology.ufl.edu/bolker/benbolker-publickey.asc