Christina Knudson (St Thomas) has kindly offered to speak about her new R package 'glmm' for fitting generalized linear mixed models. A version of this talk was presented at the useR!2017 conference.

Christina is associate professor of statistics at the University of St Thomas, Minnesota, and co-organizer of the R Ladies Twin Cities user group.

Abstract:
Conducting frequentist likelihood-based inference for generalized linear mixed models is difficult because the likelihood is a high-dimensional integral. While some R packages perform inference on an alternative function (e.g. penalized quasi likelihood), the R package glmm approximates the entire likelihood function through Monte Carlo likelihood approximation. This package conducts maximum likelihood and enables all methods of likelihood-based inference. I will give a light introduction to Monte Carlo likelihood approximation, demonstrate the use of R package glmm, describe the package's capabilities, and compare results to other packages (e.g. lme4).