[R-sig-ME] lmer for a binary dependent variable

[Andrew Beckerman]

Lillian -

There is a lengthy, satisfying, but ultimately frustration discussion  
of this issues starting with this link and the thread therein.....

https://stat.ethz.ch/pipermail/r-sig-mixed-models/2009q1/001787.html

Look in the archives under the teaching mixed models OR anything to do  
with mcmc sampling and glmm's

You might also look at Ben Bokers recent paper, and a new package  
called MCMCglmm by Jarrod Hadfield.

<http://dx.doi.org/10.1016/j.tree.2008.10.008>

Andrew
---------------------------------------------------------------------------------
Dr. Andrew Beckerman
Department of Animal and Plant Sciences, University of Sheffield,
Alfred Denny Building, Western Bank, Sheffield S10 2TN, UK
ph +44 (0)114 222 0026; fx +44 (0)114 222 0002
http://www.beckslab.staff.shef.ac.uk/

http://www.flickr.com/photos/apbeckerman/
http://www.warblefly.co.uk
----------------------------------------------------------------------------------




On 29 Jan 2009, at 14:39, Liliana Martinez wrote:

> Hi,
>
> I am trying to use the lmer function from the lme4 package in R  
> 2.8.0. to fit a generalized mixed-effects model for a dependent  
> variable with a binomial distribution (for more info on my  
> experiment, look below). However, I encounter a major problem: How  
> is it possible to find the general test statistic and see the  
> relative importance of the predictors? The methods which I found  
> described in Baayen (2008). Analyzing Linguistic Data: A Practical  
> Introduction to Statistics Using Ror on the net did not work out.  
> Here is what I got:
>
>> prec0_va2.lmer
> Generalized linear mixed model fit by the Laplace approximation
> Formula: prec_0 ~ (verb + agent)^2 + (1 | subject)
>    Data: risuvane1_binom_tolmer
>    AIC   BIC logLik deviance
>  559.7 590.5 -272.8    545.7
> Random effects:
>  Groups  Name        Variance Std.Dev.
>  subject (Intercept) 1.9975   1.4133
> Number of obs: 600, groups: subject, 30
>
> Fixed effects:
>                          Estimate Std. Error z value Pr(>|z|)
> (Intercept)                3.3120     0.5757   5.753 8.75e-09 ***
> verbzaobikaliam           -4.2031     0.5530  -7.601 2.94e-14 ***
> verbzavivam               -4.2508     0.5113  -8.313  < 2e-16 ***
> agentveh                  -2.7286     0.7219  -3.780 0.000157 ***
> verbzaobikaliam:agentveh   1.0255     0.7440   1.378 0.168058
> verbzavivam:agentveh       1.9629     0.6217   3.158 0.001591 **
> ---
> Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
>
> Correlation of Fixed Effects:
>             (Intr) vrbzbk vrbzvv agntvh vrbzb:
> verbzaobklm -0.644
> verbzavivam -0.697  0.760
> agentveh    -0.797  0.513  0.556
> vrbzbklm:gn  0.479 -0.743 -0.565 -0.491
> vrbzvvm:gnt  0.573 -0.625 -0.823 -0.584  0.620
>
>> pvals.fnc(prec0_va2.lmer)
> Error in pvals.fnc(prec0_va2.lmer) :
> mcmc sampling is not yet implemented for generalized mixed models
>
>> mcmcsamp(prec0_va2.lmer, n=500)
> Error in .local(object, n, verbose, ...) : Update not yet written
>
> Can anyone suggest a solution to this problem?
>
>
> best regards
>
> Liliana
>
>
>      _________________________________________________________
>
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models