Has anyone written an analysis in Stata for doing Bayesian analysis to see
if model A fits the data better than model B in the case where model B is
not a subset of the variables in model A but rather a different model such
as a power law versus a log-normal?