Customer Reviews

Statistical Methods in Medical Research

5 star:		(4)
4 star:		(1)
3 star:		(0)
2 star:		(1)
1 star:		(0)

 

 

This is definitely one of the most comprehensive of all biostatistics textbooks out there and it is also the best. Authoritative in style, it starts from the very basics and surveys in *detail* almost every method in biostatistics.

You will be impressed: Apart from *all* the basic techniques usually found in most biostatistics texbooks, it has two extensive chapters on Bayesian methods (with Gibbs sampling, MCMC, etc., you name it), extensive treatments of clinical trials, extensive treatments of longitudinal data and GEE models, extensive details on the bootstrap and jackknife, an extensive chapter on methods in epidemiology (risk ratios, Mantel-Haenszel method), extensive treatments of categorical data (contingency tables, logistic regression), and an extensive chapter on survival analysis. This is indeed a very extensive book.

"Statistical Methods in Medical Research" is also a book on methodology, so theorems and proofs are not to be expected. Also there are no exercises, but there are excellent illustrative examples of the various methods. While it is very descriptive, it contains enough mathematics to keep the presentation of concepts *complete*. It is also very up-to-date and has an excellent reference list.

In a nutshell, this is a work of great ambition and vision: it will cater for beginners and masters alike.

This edition if the classic text is extensively revised from previous editions. It offers a good background of theoretical considerations in biomedical statistics. It is particularly useful for the non-statistician physician who is involved in clinical trials.

As a principal level Statistician in industry, I have long used Armitage's earlier edition of this text to lend as a resource for non-statistical colleagues, because of its lucid and easily understandable explanations of statistical concepts. However, it has also proved a valuable resource to me personally, because it provided all of the formulas and calculations that I needed to manually recreate analyses; as an example, to calculate overall summary statistics from the summaries of component subgroups. I purchased this new edition for the additional content, such as Bayesian analysis, and have discovered that it still retains all of the advantages of the earlier work.

In short, I find this book to provide the most understandable introduction to statistical concepts, as well as the most useable content, of any such text that I have read. It is invaluable to both the consulting statistician and the practitioner who occasionally has to go 'outside the box' in applying statistical methods.

This is a top value product, for the student, the professionals and the specialists. This is my second buy (new edition) and I am very happy with. Really agreable to read and to use.

For a beginner, it is really an excellent book. It almost has included all of the most important statistical methods practiced in medical research without pricing at detail loss. Its way of lucid writing and rich examples can help a beginner read through the whole book with some joy. For a senior one on this field, I still think that he/she can take it as a handbook and believe he can find some clue before he pays more attention to another throughly reading.
The new edition adds a lot of new contents especially the Bayes' theroy which is becoming a fashion in nowadays.

This book is very comprehensive, so it is a good reference book at any level. However, it is often difficult to interpret and tends to jump around from topic to topic. I do not think this is a good book for someone who is looking to learn statistical methods from the ground up or for someone who is doing a self-study, but it may be useful to people who are taking introductory statistics and want a book that provides more details on methods. There are almost no examples in this book. It definitely does not hold your hand.