ABSTRACT
Developing Bayesian models requires specifying prior distributions for un-
known parameters. Central to the Bayesian philosophy is the recognition
that not only do the data, X, possess a distribution, but so do unknown
parameters, , by assumption. This idea is quite new to some researchers
and is very much at odds with the frequentist notion that unknown param-
eters are xed, unyielding quantities that can be accurately estimated with
procedures that are either repeated many times or imagined to be repeated
many times. The immobile parameter philosophy, although widespread, is
contradictory to the way that most social and behavioral scientists conduct
research. It simply is not possible to rerun elections, repeat surveys under
exactly the same conditions, replay the stock market with exactly matching
market forces, or reexpose clinical subjects to the same new stimuli.