Regression-to-the-mean modification in the MCMC

[Gates Dupont]

I'm working on an analysis that involves a regression-to-the-mean issue. I have a vague memory of a small section in one of the hierarchical modeling textbooks (I think AHM1?) about this. The section showed an example that modified the MCMC to check if the fitted relationship deviated significantly from the baked-in relationship due to regression-to-the-mean. I can't seem to find that example now, though. Does this jog anyone's memory? Sorry to be vague. Would be a big help. Thanks!!