William D. Johnson

Basic Biostatistics for Geneticists and Epidemiologists: A Practical Approach

To be an informed reader in the biomedical sciences, it is essential to have a sound grasp of the concepts related to statistics. It is less relevant that one can actually perform a particular statistical test given increasing specialization in biomedical research. The purpose of Basic Biostatistics for Geneticists and Epidemiologists: A Practical Approach is to provide the necessary framework for understanding a wide variety of common statistical tests without all the technical details needed for the calculations used in performing the tests. The challenge in meeting the objectives of this textbook is to avoid being superficial in the coverage of a statistical test without getting overly detailed in its application. In addition, the need for solid introductory explanations of the problem a particular statistical test will address is essential. One of the strengths of this text is how chapters are prefaced with careful articulation of the issue to be addressed and key concepts that are necessary to understand the problem and solutions. For example, in the chapter on significance and hypothesis testing, the authors devote 2 full pages to the principles and concepts of significance and hypothesis testing prior to embarking on any specific test. This approach leads to ageneralunderstandingof theconceptssuch that the series of actual tests become relatively easy to comprehend. This text covers the typical topics in an introductory biostatistics textbook. Chapters include descriptive statistics, probability, data distributions, variance estimators, hypothesis testing, the chi-square test, correlations and regressions, andanalysisofvariance.Perhapsthemostusefulchaptersfor the reader may be an introductory one on the role of biostatistics in biomedical research and one on populations, sampling, and study design. These chapters may be more typically found in an epidemiology textbook but provide a valuable structure from which to build an understanding of biostatistics. A chapter on advanced techniques is necessarily brief but does introduce key concepts in multivariate analysis, discriminant analysis, logistic regression, survival analysis, permutation tests, and resampling methods. The final chapter of the textbook provides a brief, concise guideline for critically evaluating published biomedical reports. This short chapter is valuable not only for those reading biomedical literature but also for us who author, edit, or review manuscripts as a reminder of good practices in communicating scientific results. There are also sections on basic genetic and epidemiology concepts. Though useful for some, these subject areas seem a bit out of place in this textbook. For those specifically interested in genetic epidemiology, more detail would be useful. For those with only a general interest in genetic epidemiology, there may be too many examples specific to genetic epidemiology. In fact, part of the title might be more appropriately labeled ‘‘in genetics and epidemiology’’ rather than ‘‘for geneticists and epidemiologists.’’ The authors make excellent use of problem sets for each chapter. Consistent with the stated focus of explaining basic concepts in biostatistics, the problems are well chosen to enhance readers’ understanding of concepts rather than computation. For those who want more depth, appendices provide more details on statistical formulations. Nonetheless, some may be intimidated by the mathematical equations in the chapters. A careful reading of the explanations of these equations in the text will guide most readers and provide greater insight into the meaning of the equations and the statistical procedure at hand. For example, the equation describing Bayes’ theorem (page 91) may seem a bit daunting. In the pages preceding this equation, however, the authors do an excellent job of developing the necessary concepts one step at a time by using carefully worded text and clear, well-explained figures that formulate each component of the theorem. By the time the full equation is presented, the reader will be comfortable with all the necessary concepts for understanding the complete theorem. Thistextbookshouldfindawideaudience.Itisappropriate as an introductory biostatistics textbookthat doesnotrequire advanced mathematics. It is particularly well suited for a course in introductory biostatistics because the early chapters provide foundations for the more advanced topics covered in the later chapters. It could also be used as statistical training for new clinical investigators who are trained physicians, particularly those interested in genetics. Laboratory scientists who want to gain more understanding of basic analyticaltechniqueswouldalsofindthistextbookquiteuseful. Anyone who wishes to critically read biomedical literature willfindtheknowledgegainedfromreadingitofgreatvalue. ACKNOWLEDGMENTS