*Review by
F. J. Rohlf (of "Biometry", Sokal & Rohlf) in Quarterly Review
of Biology 85 (1):123 (March 2010)*

The Quarterly Review of Biology, March 2010, vol. 85, no. 1

© 2010 by The University of Chicago Press. All rights reserved.

DOI: 10.1086/650289

[text repeat]

New Biological Books Miscellaneous

Analysing Cycles in Biology and Medicine: A Practical Introduction to Circular Variables and Periodic Regression. Second Edition. By K. N. I. Bell. St. John's (Canada): Razorbill Press. $25.95 (paper). xv + 163 p.; ill.; index. ISBN: 978-0-9736209-2-4. 2008.

F. James Rohlf (Ecology & Evolution, Stony Brook University, Stony Brook, New York)

This is a well-written but very introductory book on the statistical analysis of periodic data in biology. It points out the loss of information if a researcher, for example, arbitrarily divides time of year into months so that one will have groups to be analyzed using anova rather than as continuous periodic variables. The volume is amply illustrated, although many of the figures (and some of the tables) are rather busy and thus more difficult to understand. It is a short book, with just 91 pages concerned with an exposition and explanation of the methods. In addition, there are several examples shown in some detail. An appendix that provides spreadsheet formulas and Excel macros to perform the computations is also included (I would have preferred a little more space devoted to actually showing the equations involved). The third appendix, entitled Stats Refresher in a Rush, gives a 34-page overview of statistical principles. It discusses more advanced topics such as pseudo replication. Although several pages are devoted to the meaning of probability levels and the term "significance" in statistics, it would have been useful to also mention the idea of effect size. I wish the volume had included some mention of variance inflation factors (VIF). This is a more technical issue, but the possibility of large VIF values should be a concern to even beginning users of multiple regression. Unless they represent factors in a controlled experiment, independent variables are often correlated and this can generate many problems of both computation and interpretation in regression analysis.

Overall, I found the book to be a useful, interesting, and even entertaining introduction to the analysis of periodic data in biology.