Robustness of Non-Ideal Decision Procedures

Abstract

Abstract A general discussion of the robustness of decision procedures with respect to use of the wrong prior distribution emphasizes the possible usefulness of non-ideal procedures which, placing restrictions on the class of decision procedures, do not require full specification of the prior probability distribution. Applications are made to minimum mean square error estimation of a general mean and a normal variance, using non-ideal estimators linear in the sample mean and variance, respectively.