Aggregated Binary Data in MixedModels

Specific Domains Statistics
,
1

Hello all,

I’m trying to fit a generalized linear mixed model using the MixedModels package. It works fine as long as I use “normal” binary data i.e. response vector is simply 0/1 and each observartion is one row. I would like to use binary data in “aggregated” form as shown in this stackoverflow R post: https://stackoverflow.com/a/18793811 (mod2 and mod3 specification). In that case, it’s not a mixed model, but the usage is the same in R for lme4::glmer, afaik.

Is that possible in MixedModels? How?

Thanks

2

https://juliastats.org/MixedModels.jl/stable/constructors/#MixedModels.GeneralizedLinearMixedModel

The Binomial distribution is only used when the response is the fraction of trials returning a positive, in which case the number of trials must be specified as the case weights.

1 Like