As laid out here, the Bayesian median of means is a robust but always-consistent estimator of the mean (unlike e.g. the median). It applies the Bayesian bootstrap to the data to create a posterior distribution, then uses the median of the posterior. BRMS suggests using robust estimators when setting weakly-informative priors; maybe this is a good option for centering data? I might add this later on if that’s ok with you guys.
As laid out here, the Bayesian median of means is a robust but always-consistent estimator of the mean (unlike e.g. the median). It applies the Bayesian bootstrap to the data to create a posterior distribution, then uses the median of the posterior. BRMS suggests using robust estimators when setting weakly-informative priors; maybe this is a good option for centering data? I might add this later on if that’s ok with you guys.