Conditional probability as a measure of association: a Bayesian approach (with variations on the theme)

Abstract

A preference relation between two events T1 and T2 ,conditional on a given event G (usually,a statistical observation) has been introduced in previous papers by resorting to a sensible use of conditional probability. In this paper some examples are discussed to show the significance of this Bayesian approach as a measure of association of T1 and T2 with G, in contrast with a pure likelihood approach.