I'm trying to find the right way of using a mixed effects approach on some mean count data (animal visits per day to a feeder) in R. I have two interacting fixed effects (both factors) and 2 random effects (also factors, one nested in the other). While I have used a poisson GLMM on the overall data set I'm also interested in looking at mean daily visits both overall and by species. Because this produces non integer data I've tried using a Gamma glmer. This works for the mean total daily visits as there happen to be no zeros (though model is under dispersed) but all the individual species have some zeros, preventing the gamma distribution being used.

Can anybody suggest an alternative approach?

Currently trying to use a glmer in the lme4 R package, but open to using suitable alternatives.

  • $\begingroup$ You can stick with Poisson and use time as an exposure (so you can still interpret the model in terms of rates per some unit of time). $\endgroup$ – Andy W Oct 14 '15 at 17:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.