Lectures and Conferences on Mathematical Statistics and Probability
1954; Wiley; Volume: 22; Issue: 4 Linguagem: Inglês
10.2307/1907447
ISSN1468-0262
Autores Resumoa three-cornered discussion of the ideas of estimation, from the point of view of Bayes' formula, from the point of view of confidence intervals and from the point of view of fiducial argument.Since the publication of the first edition of Lectures and Conferences, there has occurred a certain shift in "allegiances" exemplified by the fact that a large section devoted to fiducial distributions, present in an early edition of a book by an eminent author, does not appear in his subsequent books, which contain, instead, sections on confidence intervals.However, indications of the confusion of the Bayes' and the more modern treatment of the problem are still noticeable in certain sections of the literature and misconceptions involved in the fiducial argument appear about as frequently.For this reason it seemed advisable to subject the matter to a detailed discussion.Here I wish to record my hearty thanks to Professor E. S. Pearson, the Editor of Biometrika, for
Referência(s)