Preface

In recent decades, longitudinal data analysis has become a topic of tremendous interest to statisticians, demographers, policymakers, and insurance companies. Scientifically, it encompasses a wide array of disciplines that include biomedical, gerontological, and behavioral and social sciences. The aim of this book is to provide a comprehensive account of the most relevant methods and techniques applied in longitudinal data analysis, without concentrating on any single field of application. Therefore, this book is expected to appeal to a wide variety of disciplines. Given my multidisciplinary background in training and research, the book covers models and methods applied in biomedicine, biostatistics, demography, psychology, sociology, and epidemiology, with practical examples associated with each of those disciplines. Linear mixed models have seen widespread applications in longitudinal data analysis, and accordingly five chapters are devoted to this statistical perspective. In light of the growing popularity of nonlinear models in empirical research, particularly in behavioral and social sciences, I also describe a variety of nonlinear mixed models in considerable detail. Regression modeling, linear and nonlinear predictions, and computing programming associated with different phases of longitudinal data analysis are extensively covered.

A large number of statistical models and methods are presented, starting with the most basic specifications and ending with some of the most advanced techniques applied in longitudinal data analysis. With a considerable volume of empirical illustrations, I attempt to make the transition from the introductory to the advanced levels as coherent and smooth as possible. For almost every major method or model, I include step-by-step instructions to demonstrate for the readers how to perform the technique on their own. Most methods described are supplemented with empirical practices, computing programs and output, and detailed interpretations of analytic results.

My hope is that scientists, professors, and other professionals of various disciplines can benefit from using this book either as a useful reference book, as a textbook in graduate courses, or any other use that requires and promotes appropriate longitudinal data modeling. This hope stems from my observation that many researchers are using incorrect models and methods to perform longitudinal analysis, particularly in handling nonlinear longitudinal data.

Given the focus on the application and practice in this book, the audience includes professionals, academics, and graduate students who have some experience in longitudinal data analyses. Numerous illustrations on various topics permit professionals to learn new models and methods or to improve their professional skills for performing longitudinal data analysis. As it covers a wide scope of methods and techniques to measure change, from the introductory to the advanced, this book is useful as a reference book for planners, researchers, and professors who are working in settings involving change. Scientists interested in studying the pattern of change over time should find it a useful guidebook for informing the appropriate models and methods for analyzing longitudinal data in their projects.

Graduate students of various disciplines constitute another important component of the audience. Social science students can benefit from the application of the concepts and the methods of longitudinal data analysis to solve problems in sociology, population studies, economics, psychology, geography, and political science. This book offers a useful framework and practical examples of applied social science, especially at a time when more questions about change are raised. The accessibility of many large-scale longitudinal datasets in the public domain will facilitate the interested student in practicing the models and methods learned from this book. Graduates students of biology, medicine, and public health, who are interested in doing research for their future careers, can learn a rich variety of techniques from this book to perform mathematical simulation, clinical trials, and competing risks analyses on health, healthcare, mortality, and disease. Longitudinal data analysis and some other related courses have been recognized as essential components of graduate students’ training in demography, psychology, epidemiology, and some of the biomedical departments. For example, in medical schools this book can bestow considerable appeal among medical students who want to know how to analyze data of a randomized controlled clinical trial adequately for understanding the effectiveness of a new medical treatment or of a new medicine on disease.

If the reader attempts to understand the entire body of the methods and techniques covered by this book, the prerequisites should include calculus, matrix algebra, and generalized linear modeling. For those not particularly familiar with this required knowledge, they might want to skip detailed mathematical and statistical descriptions and place their focus upon empirical illustrations and computer programming skills. By such a concentration, they can still command the skills to apply various models and methods effectively, thereby adding new dimensions to their professional, research, or teaching activities. Therefore, this book can be read selectively by the reader who is not extremely versed in high-level mathematics and statistics.

The reviewers for this book include Kenneth Land, Duke University; David Swanson, University of California, at Riverside; and several anonymous reviewers. Additionally, a number of colleagues and friends have enriched, supported, and refined the intellectual development of this book, including Bradley Belsher, James Edward McCarroll, Jichuan Wang, and Chu Zhang. Sincere thanks are given to Jill Tao of the SAS Institute for her help in SAS programing.

I owe special thanks to Charles C. Engel, whose consistent support and help has made the completion of this book possible. The staff of the Deployment Health Clinical Center, Defense Centers of Excellence, Walter Reed National Military Medical Center, provided tremendous dedication, competence, and excellence in the course of the preparation of this book. The assistance of Phoebe McCutchan in editing some of the graphs was vital.

Finally, I would like to thank my wife, Ming Dong, for her support and encouragement throughout the entire period of preparing, writing, and editing this book.