In a recent paper (Edwards, Lindman, and Savage, 1963), psychologists have been urged to adopt a Bayesian personal probability approach to statistical inference. The purpose of this paper is to suggest that a Bayesian logical probability approach may be superior to the personal probability approach. The concepts of frequency, personal and logical probability are defined and the characteristics and advantages of Bayesian approaches are discussed. It is shown that a theory of personal probability is dependent upon utility or preference theory. The problem of Bayes postulate is described and it is shown that an explication of the concept of logical probability is dependent upon the solution of this problem. Some work of Novick and Hall directed toward the solution of this problem is discussed. Finally, an attempt is made to integrate the concepts of personal and frequency probability within the framework of the theory of logical probability.