Longitudinal data analysis, edited by G. Fitzmaurice, M. Davidian, G. Verbeke, and G. Molenberghs (eds), Boca Raton, Chapman & Hall / CRC Press, 2009, xiv+618 pp., £49.99 or US$89.95 (hardback), ISBN 978-1-58488-658-7
This handbook might be considered as a concise but complete encyclopedia on longitudinal data analysis. The editors have made a great effort to produce a volume providing a comprehensive and up-to-date view of the theory and application of longitudinal data analysis. They are not only editors but authors or coauthors of 8 of the 23 chapters. One of the strengths of the book is the organizational structure and the fact that the book has been written by well-known experts in the field. Five different themes are covered in five different parts of the book. Each part begins with an introduction and overview of the theme treated there and briefly describes the contents of the subsequent chapters. The only exception is the first part that comprises a single well-written chapter dealing with a historical perspective on the advances made in longitudinal data analysis. In six different but cohesive chapters, the second part of this volume presents the more classical and well-known matter of the book: the parametric modelling of longitudinal data. Part three consists of five interesting chapters covering nonparametric and semiparametric methods for the analysis of longitudinal data. In these first three parts, the emphasis of the book has been on statistical models and methods for the analysis of longitudinal data with a single outcome. The next part (part four) deals with situations in which multiple outcomes, recorded simultaneously, are measured repeatedly within each subject over time. Of particular interest for some readers could be the discussion on recent developments in the analysis of high-dimensional multivariate longitudinal data. The last part of this book has seven chapters and is devoted to present alternative methods for handling missing data in longitudinal studies. In fact, the first six chapters of this last part cover the analysis of incomplete data, but the last chapter deals with the estimation of the causal effects of time-varying exposures. All the 23 chapters come with their own references facilitating the potential readers to enlarge their knowledge if necessary.
I find this book very useful for statisticians and researchers in many fields where the interest relies on studying the change of an outcome or multiple outcomes over time. Many of the chapters include examples and case studies in different disciplines and some of this material can be found in the web site of this book (http://www.biostat.harvard.edu/fitzmaur/lda). I would like to congratulate the editors and all the contributing authors for preparing this comprehensive handbook on many interesting and complementary aspects of the theory and applications of longitudinal data analysis. This handbook will have, without any doubt, an important place on the shelf of those statisticians and applied researchers working with longitudinal data.