Modern Applied Biostatistical Methods: Using S-Plus, Vol. 28

Statistical analysis typically involves applying theoretically generated techniques to the description and interpretation of collected data. In this text, theory, application and interpretation are combined to present the entire biostatistical process for a series of elementary and intermediate analytic methods. The theoretical basis for each method is discussed with a minimum of mathematics and is applied to a research data example using a computer system called S-PLUS. This system produces concrete numerical results and increases one's understanding of the fundamental concepts and methodology of statistical analysis.
Combining statistical logic, data and computer tools, the author explores such topics as random number generation, general linear models, estimation, analysis of tabular data, analysis of variance and survival analysis. The end result is a clear and complete explanation of the way statistical methods can help one gain an understanding of collected data. Modern Applied Biostatistical Methods is unlike other statistical texts, which usually deal either with theory or with applications. It integrates the two elements into a single presentation of theoretical background, data, interpretation, graphics, and implementation. This all-around approach will be particularly helpful to students in various biostatistics and advanced epidemiology courses, and will interest all researchers involved in biomedical data analysis. This text is not a computer manual, even though it makes extensive use of computer language to describe and illustrate applied statistical techniques. This makes the details of the statistical process readily accessible, providing insight into how and why a statistical method identifies the properties of sampled data. The first chapter gives a simple overview of the S-PLUS language. The subsequent chapters use this valuable statistical tool to present a variety of analytic approaches.

The book contains black-and-white illustrations.

Statistical analysis typically involves applying theoretically generated techniques to the description and interpretation of collected data. In this text, theory, application and interpretation are combined to present the entire biostatistical process for a series of elementary and intermediate analytic methods. The theoretical basis for each method is discussed with a minimum of mathematics and is applied to a research data example using a computer system called S-PLUS. This system produces concrete numerical results and increases one's understanding of the fundamental concepts and methodology of statistical analysis.
Combining statistical logic, data and computer tools, the author explores such topics as random number generation, general linear models, estimation, analysis of tabular data, analysis of variance and survival analysis. The end result is a clear and complete explanation of the way statistical methods can help one gain an understanding of collected data. Modern Applied Biostatistical Methods is unlike other statistical texts, which usually deal either with theory or with applications. It integrates the two elements into a single presentation of theoretical background, data, interpretation, graphics, and implementation. This all-around approach will be particularly helpful to students in various biostatistics and advanced epidemiology courses, and will interest all researchers involved in biomedical data analysis. This text is not a computer manual, even though it makes extensive use of computer language to describe and illustrate applied statistical techniques. This makes the details of the statistical process readily accessible, providing insight into how and why a statistical method identifies the properties of sampled data. The first chapter gives a simple overview of the S-PLUS language. The subsequent chapters use this valuable statistical tool to present a variety of analytic approaches.

Seema Sonnad

This text is an introduction to intermediate statistics. It presents statistical concepts integrated with information on how to implement them using SPLUS. A problem set and references appear at the end of each chapter. It is intended as a teaching tool for statistics. It endeavors to introduce the student simultaneously to statistical concepts and the skills and tools needed to apply them. It is intended for upper level undergraduates or graduate students taking a first course in biostatistics. The author has used the material from this text successfully in his classes for several years. It includes data sets, SPLUS syntax, and examples of outputs for each statistical concept presented. The SPLUS language is integrated into the body of the text and the book may be used, as SPLUS is, interactively. The data sets might be better provided on disk or via FTP from the author. Currently, they are long text strings that will need to be entered by hand. This book covers the topics normally found in an introductory biostatistics text: description; simulation; estimation (maximum likelihood, bootstrapping, and regression); ANOVA; and survival. It is unique in providing the instructor with a computational approach to these topics as well as their conceptual presentation. This is a major improvement over teaching concepts in class and methods in lab which are often presented with unconnected texts. I would recommend this book to any instructor teaching biostatistics courses, and especially to students interested in application. It is probably less useful outside the context of a class, except for those motivated to teach themselves from the ground up or those looking for SPLUS codes toimplement statistical concepts learned in an introductory class.

Steve Selvin, Ph.D., is Professor of Biostatistics and Epidemiology at the University of California School of Public Health, Berkeley. He is also the author of Statistical Analysis of Epidemiologic Data, 2nd ed. (OUP, 1996).