Excerpts

Statistics for the Life Sciences is an introductory text in statistics, specifically addressed to students specializing in the life sciences. Its primary aims are (1) to show students how statistical reasoning is used in biological, medical, and agricultural research; (2) to enable students confidently to carry out simple statistical analyses and to interpret the results; and (3) to raise students' awareness of basic statistical issues such as randomization, confounding, and the role of independent replication. Style and Approach The style of Statistics for the Life Sciences is informal and uses only minimal mathematical notation. There are no prerequisites except elementary algebra; anyone who can read a biology or chemistry textbook can read this text. It is suitable for use by graduate or undergraduate students in biology, agronomy, medical and health sciences, nutrition, pharmacy, animal science, physical education, forestry, and other life sciences. Use of Real Data. Real examples are more interesting and often more enlightening than artificial ones. Statistics for the Life Sciences includes hundreds of examples and exercises that use real data, representing a wide variety of research in the life sciences. Each example has been chosen to illustrate a particular statistical issue. The exercises have been designed to reduce computational effort and focus students' attention on concepts and interpretations. Emphasis on Ideas. The text emphasizes statistical ideas rather than computations or mathematical formulations. Probability theory is included only to support statistics concepts. Throughout the discussion of descriptive and inferential statistics, interpretation is stressed. By means of salient examples, the student is shown why it is important that an analysis be appropriate for the research question to be answered, for the statistical design of the study, and for the nature of the underlying distributions. The student is warned against the common blunder of confusing statistical nonsignificance with practical insignificance, and is encouraged to use confidence intervals to assess the magnitude of an effect. The student is led to recognize the impact on real research of design concepts such as random sampling, randomization, efficiency, and the control of extraneous variation by blocking or adjustment. Numerous exercises amplify and reinforce the student's grasp of these ideas. The Role of the Computer/ The analysis of research data is usually carried out with the aid of a computer. Computer-generated graphs and output, either from the statistical software DataDesk or MINITAB, are shown at several places in the text. MINITAB commands are given in a number of places (although MINITAB output can also be generated from menus while running the software). However, in studying statistics it is desirable for the student to gain experience working directly with data, using paper and pencil and a hand-held calculator, as well as a computer. This experience will help the student appreciate the nature and purpose of the statistical computations. The student is thus prepared to make intelligent use of the computer--to give it appropriate instructions and properly interpret the output. Accordingly, most of the exercises in this text are intended for hand calculation. Selected exercises, identified with the words "computer exercise" are intended to be completed with use of a computer. (Typically, the computer exercises require calculations that would be unduly burdensome if carried out by hand.) Organization This text is organized to permit coverage in one semester of the maximum number of important statistical ideas, including power, multiple inference, and the basic principles of design. By including or excluding optional sections, the instructor can also use the text for a one-quarter course or a two-quarter course. It is suitable for a terminal course or for the first course of a sequence. The following is a brief outline of the text: Chapter 1: Introduction. The nature and impact of variability in biological data. Chapter 2: Orientation. Frequency distributions, descriptive statistics, the concept of population versus sample. Chapters 3, 4, and 5: Theoretical preparation. Probability, binomial and normal distributions, sampling distributions. Chapter 6: Confidence interval for a mean or for a proportion. Chapter 7: Comparison of two independent samples. The t -test and the Wilcoxon-Mann-Whitney test. Chapter 8: Design. Randomization, blocking, hazards of observational studies. Chapter 9: Inference for paired samples. Confidence interval, t-test, sign test, and Wilcoxon signed-rank test. Chapter 10: Categorical data. Chi-square goodness-of-fit test, conditional probability, contingency tables. Optional sections cover Fisher's exact test, McNemar's test, and odds ratios. Chapter 11: Analysis of variance: one-way layout. Multiple comparison procedures, two-way analysis of variance, contrasts, and interaction in two-factor designs are included in optional sections. Chapter 12: Regression and correlation. Descriptive and inferential aspects of simple linear regression and correlation and the relationship between them. Chapter 13: A summary of inference methods. Statistical tables are provided at the back of the book. The tables of critical values are especially easy to use, because they follow mutually consistent layouts and so are used in essentially the same way. Optional appendices at the back of the book give the interested student a deeper look into such matters as how the Wilcoxon-Mann-Whitney null distribution is calculated. Excerpted from Statistics for the Life Sciences by Myra L. Samuels, Jeffrey A. Witmer All rights reserved by the original copyright owners. Excerpts are provided for display purposes only and may not be reproduced, reprinted or distributed without the written permission of the publisher.