|
| Author(s) |
Bromek, T. (ed.) Plesszczynska, E. (ed.) |
|---|---|
| Title | Statistical Inference, Theory and Practice |
| Publisher | Kluwer Academic Publishers |
| Year of publication | 1991 |
| Reviewed by | Teodor TOADERE |
This book deals with selected ideas and schemes of statistical inference which are of considerable value in practice. Stress has been laid upon problems of observability and of reducing the information on the empirical phenomena investigated. The whole subject has being treated in a rather unconventional manner.
The book contains 11 chapters, which are grouped in three parts. In Part One (two chapters) the consecutive stages of formalization of statistical problems are presented, i.e., the description of the experiment, the presentation of the aim of the investigation and of the constraints put upon the decision rules. The authors prove that the links between the real problem and its formal counterpart are often so weak that the solution of the formal problem may have no rational interpretation at the practical level.
The Part Two (three chapters) presents some selected theoretical topics, namely discriminant analysis, screening and measuring stochastic dependence and it is emphasized their interrelations. But only the practical problems discussed in Part Three (six chapters) make it plain to the reader what are the essential difficulties which have to be tackled by statistics.
Contributor at this book are: T. Bromek, J. Cwik, W. Gafrikova, M. Golembiewska-Skoczylas, K. Grygiel, T. Kowalczyk, J. Mielniczuk, M. Niewiadomska-Bugaj, E. Pleszczynska, W. Szczesny, H. Szczotka, Z. Szczotkowa. Though the list of co-authors contains several names, the book presents a unified style.
The book also contains: Closing remarks, one Appendix with Algorithms for evaluating monotone dependence function and screening threshold, a rich list of references and an index.
The material found in this bookie meant for professional statisticians as well as for people interested in investigating empirical phenomena, and will contribute materially to the crossdisciplinary appreciation of statistical science.