Biostatistical Analysis 5th Edition

Book Preface

Beginning with the first edition of this book, the goal has been to introduce a broad array of techniques for the examination and analysis of a wide variety of data that may be encountered in diverse areas of biological studies. As such, the book has been called upon to fulfill two purposes. First, it has served as an introductory textbook, assuming no prior knowledge of statistics. Second, it has functioned as a reference work consulted long after formal instruction has ended.

Colleges and universities have long offered an assortment of introductory statistics courses. Some of these courses are without concentration on a particular field in which quantitative data might be collected (and often emphasize mathematics and statistical theory), and some focus on statistical methods of utility to a specific field (such as this book, which has an explicit orientation to the biological sciences). Walker (I 929: 148-163) reported that, although the-teaching of probability has a much longer history, the first statistics course at a\~.S. university or college probably was at Columbia College (renamed Columbia University in 1896) in 1880 in the economics department; followed in 1887 by the second-the first in psychology-at the University of Pennsylvania; in 1889 by the first in anthropology, at Clark University; in 1897 by the first in biology, at Harvard University; in 1898 by the first in mathematics, at the University of Illinois; and in 1900 by the first in education, at Teachers College, Columbia University. In biology, the first courses with statistical content were probably taught by Charles B. Davenport at Harvard (1887 -1899), and his Statistical Methods in Biological Variation, first published in 1899, may have been the first American book focused on statistics (ibid.: 159).

The material in this book requires no mathematical competence beyond very elementary algebra, although the discussions include many topics that appear seldom, if at all, in other general texts. Some statistical procedures are mentioned though not recommended. This is done for the benefit of readers who may encounter them in research reports or computer software.

Many literature references and footnotes are given throughout most chapters, to provide support for material discussed, to provide historical points, or to direct the reader to sources of additional information. More references are given for controversial and lesser-known topics.

The data in the examples and exercises are largely fictional, though generally realistic, and are intended to demonstrate statistical procedures, not to present actual research conclusions. The exercises at the end of chapters can serve as additional examples of statistical methods, and the answers are given at the back of the book.

The sample sizes of most examples and exercises are small in order to conserve space and to enhance the ease of presentation and computation. Although the examples and exercises represent a variety of areas within the biological sciences, they are intended to be understood by biology students and researchers across a diversity of fields.

There are important statistical procedures that involve computations so demanding that they preclude practical execution without appropriate computer software. Basic principles and aspects of the underlying calculations are presented to show how results may be obtained; for even if laborious calculations will be performed by computer, the biologist should be informed enough to interpret properly the computational results. Many statistical packages are available, commercially or otherwise, addressing various subsets of the procedures in this book; but no single package is promoted herein.

A final contribution toward achieving a book with self-sufficiency for most biostatistical needs is the inclusion of a comprehensive set of statistical tables, more extensive than those found in similar texts.