I have a response variable of proportions representing the number of hours an individual is performing a specific activity divided by the total number of hours in the study period. I have no 0's or 1's in this data set. I am interested in knowing the extent to which specific factors (such as season) influence this proportion. I have random effects (individual ID) and fixed effects, and so I am considering a generalized linear mixed model.

I have read that the 'glmer' function can be used with a 'binomial' family and a 'weights' argument. The following is from the help file of glm:

The help file of glm says:

 "For a binomial GLM prior weights are used to give the number of trials 
  when the response is the proportion of successes" 

I am not sure what to specify for the 'weights' argument in my case, since my proportions are not representative of successes/total or failure/total.

Other sources state that, if data is "inherently proportional" (which I would think my data is), then a beta regression is more suitable. I do not think I can incorporate random effects with a beta regression. My question is thus: which approach would be most suitable for my data?


1 Answer 1


You could use glmmTMB to do beta regression using the argument family=beta_family. It can do either GLMs or GLMMs, so random effect can be included. I don't think the beta family has been extensively tested, but it should work as far as we know.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.