ABSTRACT
Significance tests and confidence intervals constitute the central topics of most chapters in this volume, with the tacit suggestion that there are no other mechanisms with which to make inductive statistical inferences. It would be a pity for the reader to believe this, however, because one of the most respected of all modes of inductive statistical reasoning makes no reference whatsoever to either significance tests or confidence intervals. As can be seen from my title, I have in mind work done to advance Bayesian statistical methods. Bayesian and classical forms of statistical inference are fundamentally different from one another be-
cause these systems entail different assumptions at the outset. I first examine basic concepts associated with one of the simplest classical inferential problems and then go on to describe its Bayesian or subjectivist counterpart. Specifically, in the next three paragraphs when terms like probability and parameter are used, the context intended is that of classical or frequentist statistical inference; but for virtually all of the remainder of this chapter these and other related technical terms are used with reference to the Bayesian or subjectivist paradigm.
