Simultaneous estimation of poisson means under exchangeable and independence models

Abstract

Consider two problems: (1) the simultaneous estimation of a number of Poisson parameters when the parameters are believed a priori to be exchangeable and (2) the simultaneous estimation of Poisson parameters in a two-way contingency table where the classification variables are believed independent. By the use of hierarchical prior distributions, simple new point and interval estimates are proposed for the parameters of interest. The methods are applied on home run data of Ted Williams and Joe DiMaggio.