Caesarean section rates differ dramatically in different regions of the United States, apparently a result of medical students and new doctors learning by imitation and adopting the customary practices of their workplaces. Given information about a doctor’s medical school and residency locations, one can predict her views on appropriate circumstances for C-sections, i.e. her actual opinion is less relevant in Bayesian calculations, since it reflects a known ‘meme lineage.’
I just commented at http://www.typepad.com/t/co... dthat Carl is right, we need to take into account correlations between errors when estimating a best average consensus belief.
I just commented at http://www.typepad.com/t/co... dthat Carl is right, we need to take into account correlations between errors when estimating a best average consensus belief.