Statistical information and likelihood. A collection of critical essays by Dr. D. Basu (Q1188506)

These essays are a sample of \textit{D. Basu}'s statistical work from roughly 1955 to 1982. Almost all of them have been reviewed on first publication (though this collection presents some of these papers in slightly different form to the original), and it seems unnecessary to review the papers individually here. The essays are collected in three groups: I, Information and likelihood; II, Survey sampling and randomization; and III, Miscellaneous notes and discussions. Many are preceded by notes by the editor and by the author, those of the former being limited mainly to details on the publication or first presentation of the paper, while those of the latter provide useful additions to the essays. The papers in the first part of the book deal with the recovery of ancillary information, statistical information and likelihood, sufficiency and nuisance parameters. In the second part, sufficiency and likelihood in survey sampling, along with randomization, are considered, while the third part contains essays on likelihood, the Fisher exact test, unbiased estimation, asymptotic efficiency and the Basu theorems. In a sense the essays describe the tentative steps of a passage to Bayesian enlightenment. As such a record the monograph provides a useful account of a perhaps slightly reluctant conversion to Bayesianism, and it should be on the shelves of all who are interested in this topic.