Observational Studies

An observational study is an empirical investigation of the effects of treatments, policies or exposures. They differ from experiments in that the investigator cannot control the assignment of treatment to subjects. Scientists across a wide range of disciplines undertake such studies and the aim of this text is to provide a sound statistical account of the principles and methods for the design and analysis of observational studies. Readers are assumed to have a working knowledge of basic probability and statistics, but otherwise the account is reasonably self-contained. Throughout, there are extended discussions of actual observational studies to illustrate the ideas discussed. These are drawn from topics as diverse as smoking and lung cancer, lead in children, nuclear weapons testing, and placement programmes for students. As a result, many researchers involved in observational studies should find this a useful companion to their work.

From the reviews of the second edition:

TECHNOMETRICS

"This is certainly a book that persons involved with studying the effects of treatments or policies on the affected subjects will want to own. It would seem that persons would want to replace their 1E with this updated and considerably expanded edition."

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION

"Rosenbaum provides a clear presentation of concepts and language (e.g., overt vs. hidden bias, matching, referent as opposed to control subjects), and presents specific methodological procedures as well as larger philosophical issues...an important book that should be widely read."

D. OAKES, UNIVERSITY OF ROCHESTER, USA

"This is a fascinating book which combines elegant theory with good practical advice on applications, including in-depth discussion of many interesting examples from diverse fields."  

SHORT BOOK REVIEWS

"This second edition is about fifty percent larger than the original. The flavour is the same. The book provides a thoughtful discussion, at a mathematical and conceptual level, of what can and cannot be learned from observational studies."

"There are many new examples and problems which enrich the reading of the book. The structure of having problems and bibliographic notes is very useful. ... There is a very thorough study of hidden bias ... which makes this book very different from many existing books in epidemiology." (Anup Dewanji, Sankhya, Vol. 65 (3), 2003)

"The book by Paul R. Rosenbaum is devoted to the investigation of such observational studies which explore the effects of exposures, treatments, or policies but which are characterized by the fact that the investigator lacks control of treatment assignment. Rosenbaum gives a concise and mathematically founded approach to a careful interpretation of the findings of observational studies ... . The book becomes especially valuable for applied researchers by numerous carefully discussed examples from different fields of application, which supplement the given theoretical arguments ... ." (I. Pigeot, Metrika, August, 2003)

"The purpose of this book is to give an account of statistical principles and methods for the design and analysis of observational studies. ... I believe this book will be very useful to researchers and graduate students in the biomedical and social sciences. The distinction between randomized experiments and observational studies is of vital importance for the difficult task of providing correct conclusions from observational studies. Without any doubt, I would recommend this new edition be added to our libraries as an interesting reference book." (M. Dolores Ugarte, Journal of Applied Statistics, Vol. 30 (10), 2003)

"The second edition of this remarkable book is about fifty percent larger than the first, published in 1995. ... Technical issues are treated thoroughly but do not interfere with conceptual matters. Each chapter concludes with problems both conceptual and technical. The mathematical level is variable. The book would be adequate for a seminar course for students with some previous knowledge of the subject area." (Ricardo Maronna, Statistical Papers, Vol. 45 (1), 2004)

"This second edition of a well-known and useful book has been expanded and revised ... . Observational studies, whether retrospective or prospective, represent the majority of research in the medical, social, and environmental sciences and are very different from designed experiments. Rosenbaum provides a clear presentation of concepts and language (e.g., overt vs. hidden bias, matching, referent as opposed to control subjects), and presents specific methodological procedures as well as larger philosophical issues. ... Observational Studies is an important book that should be widely read." (Peter C. Wollan, JASA, December, 2004)

"Here is a book in which the author has made a considerable effort in producing a second edition (2E) for a specialized book. He has turned a slender little book ... into a substantial and comprehensive text. ... This is certainly a book that persons involved with studying the effects of treatments or policies on the affected subjects will want to own. It would seem that persons would want to replace their 1E with this updated and considerably expanded edition." (Technometrics, Vol. 45 (2), May, 2003)

"An observational study is an empirical investigation of the effects caused by a treatment, policy, or intervention, in which it is not possible to assign subjects at random to treatments or controls, as would be done in controlled experiments. ... This book is extremely useful to researchers and graduate students in the biomedical and social sciences." (T. Postelnicu, Zentralblatt MATH, Vol. 985, 2002)