Applied Longitudinal Analysis: (Wiley Series in Probability and Statistics)

Praise for the First Edition

". . . [this book] should be on the shelf of everyone interested in . . . longitudinal data analysis."
—Journal of the American Statistical Association

Features newly developed topics and applications of the analysis of longitudinal data

Applied Longitudinal Analysis, Second Edition presents modern methods for analyzing data from longitudinal studies and now features the latest state-of-the-art techniques. The book emphasizes practical, rather than theoretical, aspects of methods for the analysis of diverse types of longitudinal data that can be applied across various fields of study, from the health and medical sciences to the social and behavioral sciences.

The authors incorporate their extensive academic and research experience along with various updates that have been made in response to reader feedback. The Second Edition features six newly added chapters that explore topics currently evolving in the field, including:

• Fixed effects and mixed effects models
• Marginal models and generalized estimating equations
• Approximate methods for generalized linear mixed effects models
• Multiple imputation and inverse probability weighted methods
• Smoothing methods for longitudinal data
• Sample size and power

Each chapter presents methods in the setting of applications to data sets drawn from the health sciences. New problem sets have been added to many chapters, and a related website features sample programs and computer output using SAS, Stata, and R, as well as data sets and supplemental slides to facilitate a complete understanding of the material.

With its strong emphasis on multidisciplinary applications and the interpretation of results, Applied Longitudinal Analysis, Second Edition is an excellent book for courses on statistics in the health and medical sciences at the upper-undergraduate and graduate levels. The book also serves as a valuable reference for researchers and professionals in the medical, public health, and pharmaceutical fields as well as those in social and behavioral sciences who would like to learn more about analyzing longitudinal data.

Preface to First Edition xxi

Acknowledgments xxv

Part I. Introduction to Longitudinal and Clustered Data

1. Longitudinal and Clustered Data 1

2. Longitudinal Data. Basic Concepts 19

Part II. Linear Models for Longitudinal Continuous Data

3. Overview of Linear Models for Longitudinal Data 49

4. Estimation and Statistical Inference 89

5. Modelling the Mean: Analyzing Response Profiles 105

6. Modelling the Mean: Parametric Curves 143

7. Modelling the Covariance 165

8. Linear Mixed Effect Models 189

9. Fixed Effects versus Random Effects Models 241

10. Residual Analyses and Diagnostics 265

Part III. Generalized Linear Models for Longitudinal Data

11. Review of Generalized Linear Models 291

12. Marginal Models: Introduction and Overview 341

13. Marginal Models: Generalized Estimating Equations (GEE) 353

14. Generalized Linear Mixed Effects Models 395

15. Generalized Linear Mixed Effects Models: Approximate Methods of Estimation 441

16. Contrasting Marginal and Mixed Effects Models 473

Part IV. Missing Data and Dropout

17. Missing Data and Dropout: Overview of Concepts and Methods 489

18. Missing Data and Dropout: Multiple Imputation and Weighting Methods 515

Part V. Advanced Topics for Longitudinal and Clustered Data

19. Smoothing Longitudinal Data: Semiparametric Regression Models 553

20. Sample Size and Power 581

21. Repeated Measures and Related Designs 611

22. Multilevel Models 627

Appendix A. Gentle Introduction to Vectors and Matrices 655

Appendix B. Properties of Expectations and Variance 665

Appendix C. Critical Points for a 50:50 Mixture of Chi-Squared Distributions 669

References 671

Index 695

Nan M. Laird, PhD, is Professor of Biostatistics at the Harvard School of Public Health. A Fellow of the American Statistical Association and Institute of Mathematical Sciences, she has published extensively in the areas of statistical genetics, longitudinal studies, missing or incomplete data, and analysis of multiple informant data.

James H. Ware, PhD, is Frederick Mosteller Professor of Biostatistics at the Harvard School of Public Health. A Fellow of the American Statistical Association and statistical consultant to the New England Journal of Medicine, he has made significant contributions to the development of statistical methods for the design and analysis of longitudinal studies.

“The text is well-organized and clearly written. It is accessible to researchers with varying levels of statistical expertise, with plenty of data examples that make reading and learning enjoyable. I recommend it to biostatisticians as well as to clinicians and other health researchers who may not have much statistical training . . . Applied Longitudinal Analysisis generally my first recommendation when asked for a valuable resource in the field due to the breadth of topics covered and its practical utility.”  (Journal of Biopharmaceutical Statistics, 1 January 2013)

“The book also serves as a valuable reference for researchers and professionals in the medical, public health, and pharmaceutical fields, as well as those in social and behavioral sciences who would like to learn more about analysing longitudinal data.” (Zentralblatt MATH, 2012)