Comparing Methods of Parameter Estimation for Fitting Models of Clustered Binary Responses

We found that the most accurate parameter estimation methods for dealing with clustered binary data in GLMMs were those in the lme4 and NLMIXED packages.

For my undergraduate honors thesis, I worked with Dr. Xianming Tan on comparing methods of parameter estimation for fitting models of clustered binary responses, primarily through GLMM and GEE approaches.