ABSTRACT
We now turn to situations in which all members of the family of distributions are to be treated as of equal importance. Provisionally we assume that one of the distributions in the family is the true dis tribution, an assumption which is, o f course, always tentative. What can we say about the true parameter value? This can reasonably be regarded as the central problem of statistical inference, and the results of Chapters 4-6, while important in their own right, are in a sense a preliminary to the present Chapter.
