A book which is reliable to read, especially for reserachers, post-grad students

I used this book for my psychological statistics course last semester. Although I have had teachers in the past who explained things more adequately, this book did a really good job explaining things for the most part. I almost always use examples to learn how to apply knowledge, as working through a problem step by step seems to be the best teacher to me. Fortunately, this book has a lot of examples. And, as a bonus, these examples apply towards the social sciences so you can see how everything fits in with a topic you're actually interested in (let's face it, few are interested in math).

I had already taken a statistics course in high school, so my stats class last year was pretty much a review for me. However, I wound up actually learning more than I planned to because of this book.

Overall, I recommend the book for anybody learning statistics.

Modern statistics programs in universities rightly give heavy weight to the development of technical ability and understanding. However, many of those graduating from such programs need inter-personal skills to communicate with scientists who are not statisticians. Dr. Derr's book and accompanying video help the reader to develop such skills in a series of meaningful exercises.

The book is also designed for use as in a graduate statistics course in consulting. Statistics departments in universities that do not yet have such courses should start one, and will find this text an admirable basis for such an endeavor.

Estimating functions or curves to describe data is fundamental to data analysis and is very practical. There are however, many deep philosophical issues. Tarter is a Professor of Statistics at Berkeley. He has published articles and an earlier book on nonparametric curve estimation. This book is unique as it is very philosophical and presents many of the old ideas from the time of Fisher, Neyman and Karl Pearson. The purpose is to reconsider the thinking of these geniuses of the early 20th century in light of recent advances in nonparametric and computer-intensive methods. Much of the material came from a biostatistics class he taught at the University of Washington in Seattle during the summer of 1970.

This book is intended for graduate level students in statistics.

This book is succint, 187 pages excluding the bibliography. It starts from scratch (no statistical background) introduces the relative frequency and Bayesian (subjective) interpretations of probability. Both viewpoints are given throughout the book, where appropriate. The last chapter is a very useful treatment of deconvolution in the presence of limited resolution of a measuring device, and it makes the connection to image restoration.

Time series analysis is not included.

The author's background is particle physics, and he works at CERN. Many examples come from that arena, but the presentation is definitely more accessible to an engineer, physicist or chemist than the numerous treatments by financial analysts and social scientists.

In this volume Federer provides a variety of statistical designs for intercropping experiments when three or more crops are involved. He describes a variety of different situations and provides specialized solutions. Topics include: (1) constant densities (Chapter 13) , (2) varying densities (Chapter 14), (3) effects of mixing when individual cultivar responses are available (Chapter 15), (4) mixtures when individual responses are not available (Chapter 16)and more. This is the second volume of a text on intercropping experiments and anyone with enough interest to read volume II should certainly get volume I as well. It is well-written, well-researched and presented by an expert in the field of design of agricultural experiments.

This book presents better ways to maximize the information one gets from an intercropping experiment. As the author states the difficulties with three or more crops is an order of magnitude greater than with only two crops. Nevertheless such experiments are common and in today's computing environment the difficulties can be overcome. Due to the algebraic complexity Federer used symbolic computing through tools such as GAUSS, MATHEMATICA and MAPLE to obtain trustworthy results.

Federer has prepared a bibliography of over 3000 references which is so large that it is not included in the text volumes. It can be obtained by going to his university web site or in hard copy or ordered on diskette This is discussed in Chapter 20. The text does contain many other references including research articles and related books in the literature cited sections of the various chapters.

This is a reprint of Peter John classic book on experimental designs originally published by MacMillan but now reprinted by the Society for Industrial and Applied Mathematics in their Classics in Applied Mathematics series. The book was written in 1971 and there have been many changes in philosophy for experimental designs since then. Also the revolution in computing has had an influence. In his "Preface to the Classics Edition" John reflects on the state of computing in 1966 when he first contracted to write the book and 1998 when the book was reprinted. He also describes the history of experimental design and extends it to the advances of Taguchi. It was industrial applications that attracted John to do research in experimental designs and he see industry as a rich source for design problems. The text covers all the classical design work with examples from agriculture and industry. It has heavy coverage of incomplete block designs and in Chapter 13 he demonstrates how to construct the various balanced incomplete block designs. There is a lot of theory. Although the title says design and analysis it is primarily a book on designs. Response surface methods and the evolutionary operation approach are also covered....

Presents a variaty of subjects in a concise and well defined manner. It unifies many methods like PCA, factor analysis, arma modeling and others. It is best suited to people with good math background for which too many words without exact definitions means nothing. I found this a problem in many factor analysis texts. Its only disadvantage is that the terminology is slightly different than the usual one. Overall, it is worth every cent of it !

After desperately searching for a good book in statistical estimation theory in the library I found this book to be an excellent resource and worth buying. It has easy explanations and the examples help one better understand the material. It covers both estimation and detection theory. This book is probably best for the engineering student that is more concern with applying the concepts than deriving theorems.

This book gives a good introduction to bayesian statistics, especially the relation between bayesian and frequentistic statistics is well explained. Together with "The Bayesian Choice" and "Data Analysis - A Bayesian Tutorial" it covers my initial need for knowledge about Bayesian statistics.