NIOSHTIC-2 Publications Search

Statistics with applications in biology and geology (Book Review).

Technometrics 2004 Feb; 46(1):111
If teaching statistics by example is your goal, then this book will provide a wealth of interesting examples that demonstrate the methods of statistics and practical data analysis. The focus of the book is biology and geology, as the title indicates, with a greater emphasis on biology. There are about 70 separate datasets in the examples and exercises, providing a great deal of material for teaching. About one-third ofthe examples and exercises are based on problems in geology, including sediment composition, sound propagation in rock, metal content of geologic samples, sand sifting rates, pH of core samples, shape and weight of stones, magnetism in lava flows, gas diffusion in stones, mineral content in water, earthquakes, sediment transport, wind direction, and orientation of crystals. The biological exercises and examples encompass problems related to fisheries taxonomy, vertebrate physiology, invertebrate population density and size distribution, ecology, toxicology, both plant and animal genetics, agronomy, microbiology, and human studies involving asthma, twin studies, cancer risks, chronic disease, and physiology. The authors provide the data on their website along with SAS programs used for analysis. The book devotes more pages to displaying SAS program code than I enjoy reading, and much of this is redundant with the website, but the duplication of material may be helpful for students. The first example of a SAS program includes a SAS macro and programming using PROC IML, a matrix programming language within SAS. This is much too advanced for a student's first encounter with SAS, and is a challenge to even experienced SAS programmers. However, the website is well organized and enables the user to easily locate the programs for the specific examples. . The book comprises 12 chapters, beginning with a discussion of statistical models and inference in Chapter 1. Chapter 2 emphasizes graphics including histograms and probability plots. Chapter 3 discusses normally distributed data procedures, including one- and two-sample tests, one-way analysis of variance, and simple regression. Chapter 4 presents two-way analysis of variance with interaction as an example that extends the simple linear models. Chapter 5 discusses the concepts of power, non central distributions, and sample size efficiency through experimental design. Chapter 6 introduces correlation and the bivariate normal distribution. Chapter 7 presents the multinomial distribution and statistical methods for categorical data, and Chapter 8 introduces the Poisson distribution for analysis of rate data. Chapter 9 introduces Poisson and logistic regression as part of a brief discussion of generalized linear models. Chapter 10 presents methods for analyzing directional data in two and three dimensions using circular normal distributions (von Mises and Fisher distributions). Although concepts of maximum likelihood are briefly mentioned in various parts of the book, Chapter 11, 'The Likelihood Method," offers more details and a discussion of quadratic approximation. Chapter 12 introduces some nonparametric tests for one-sample, two-sample, and k-sample KruskalWallis tests. Unfortunately, my copy of the book had too many printing errors, especially in the mathematical notation, which would cause confusion for students seeing this material for the first time. Some of the discussion assumes a background that is too advanced for the intended audience of undergraduate students in biology and geology. Discussions of subspaces, projections about the geometry of linear models, and the distinctions between "affine subspace" and "linear subspace" in the section on generalized linear models seem to be a distraction from the theme of practical data analysis in biology and geology. The book's strength lies in the emphasis on practical examples and realistic research problems that include many interesting datasets. The analyses are presented in a way to facilitate teaching the principles of statistics and the importance of quantitative studies to students in the natural sciences.
Statistical-analysis; Mathematical-models; Data-processing; Biological-factors; Geology; Teaching; Education
Publication Date
Document Type
Journal Article
Fiscal Year
Issue of Publication
NIOSH Division
Priority Area
Source Name
Page last reviewed: March 11, 2019
Content source: National Institute for Occupational Safety and Health Education and Information Division