[R-sig-ME] Mixed effect logistic regression help

[Christine Griffiths]

Dear Giancarlo,

Thank you for very much for your advice. I suspected that may be the case. 
Do you know if the Zelig package 
(<http://cran.r-project.org/web/packages/Zelig/vignettes/logit.mixed.pdf>) 
can accommodate pairing, or is simply accounting for the variance?

Many thanks in advance,
Christine

--On 05 April 2011 04:54 +0000 giancarlo <gsadoti at syr.edu> wrote:

> CJ Griffiths <Christine.Griffiths at ...> writes:
>
>>
>> Dear all,
>>
>> I want to specify a mixed conditional logistic regression to model
>> microhabitat selection, but am unsure whether my dataframe and model are
>> correct. I want to compare parameters such as wind and temperature at the
>> location of the animal (1) to a random observation, where the animal was
>> absent (0). Each Response (1/0) is thus paired by the variable Micro. To
>> account for this pairing, I specified the random effect as 1|Micro.
>> However, I repeatedly sampled 11 animals (Ind). Random effect = 1|Ind
> ...
>
> A little late in the response...
>
> To my understanding, mixed models cannot accommodate the paired 1/0
> structure of your data.  Adding micro as a mixed effect is only
> accounting for the variance in your 1/0 response among different levels
> of micro.  However, since you have an equal number of 1s and 0s for each
> level, you see zero as the variance in micro.  You need to use a
> conditional logistic model (e.g., clogit in package survival) to do what
> you want to do to accommodate the paired case/control nature of your
> samples.  However, clogit cannot accommodate the repeated sampling of
> individuals (which lmer, nlme, etc. can, but not in this case).  The only
> real application I've seen bridging these two 'worlds' is the following:
>
> Duchesne, T., D. Fortin, and N. Courbin. 2010. Mixed conditional logistic
> regression for habitat selection studies. Journal of Animal Ecology
> 79:548-555.
>
> Unfortunately, it was implemented in Matlab, so not so useful to any of
> us in R land.  Personally, I'd be very excited to see a mixed conditional
> logit package or extension developed for R, but I'm neither statistician
> nor programmer.
>
> Giancarlo
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models



----------------------
Christine Griffiths
PhD student
School of Biological Sciences
University of Bristol
Woodland Road
Bristol BS8 1UG
Tel: 0117 9288918
Fax 0117 3317985
Christine.Griffiths at bristol.ac.uk
http://www.bristol.ac.uk/biology/person/index.html?personKey=jBoLZvRjH3he6kxfWwIXzKqzcOtvUW