Replicate Stata 'variance' and 'robust' in R for mixed effects methods [closed]

Question

I am working with a small dataset of clustered data. The experiment is on mice, testing the efficacy of a substance for neurons cells within mice, and delivery of a fatty acid substance between mice. This results in interesting within and between effects.

The authors have published their results (article, github)

However, their results were from the outputs provided in Stata. When reproducing in R (or SAS), I am getting very different results for the standard errors.

Using lme in R, is there a clearer way of reproducing the robust standard errors and variance options provided by in Stata?

Answer 1

There are various versions of robust standard errors, e.g., using the sandwich estimator, (grouped) jackknife, the Bootstrap method. It is not clear which one you are looking for.

The GLMMadaptive package implements the sandwich estimator for mixed models; for more info check here.