[Home] [Up]

My Own Review

What I Told

Oxford University Press

The book provides a thorough treatment of analysis of variance, integrating issues of experimental design, in some 250 pages.  A virtually notation-free presentation is offered.  Topics not often discussed in texts at this level are covered, including measurement issues and missing data.  The chapter on functional measurement is unique.  Quite a bit of the material is original and has previously appeared only in journals.  There are exercises, with answers, for all of the chapters.  Every technique introduced has one or more practice problems, with the main analyses having several.  Graphing the data is emphasized, and tables and graphs are included within the answers.  The book is fully indexed, and includes a glossary of terms from introductory statistics.

It is very much a “how-to” book, with the emphasis on setting up the experiment and analyzing and presenting the data.  There is little attention devoted to mathematical underpinnings.  The target audience is psychologists, not mathematical statisticians.  While it is unlikely that a student will consider any statistics book pleasure reading, this one is by far the simplest to read.

The accompanying computer package is an easy-to-use, yet semi-professional quality, set of ten WINDOWS programs.  The programs provide graphs as well as tables.   The programs have help files in addition to on-screen instructions, and can optionally print directly to WORD.  The opening screens of all the computational programs ask the user about the nature of the design, with the questions phrased in English.  Data entry is similar for all of the programs, and is simplified by the inclusion of factorial indices within the grid presented for data entry.  The text contains screen shots illustrating program features.

I used the combination for several years in my senior undergraduate and introductory graduate courses in statistics in the psychology department.  A student typically learns to use any of the programs within five minutes, and that knowledge transfers so that other programs can be used at first sight.  The programs have been tested by several generations of students, and praise is universal (especially among those who have previously been tortured by SPSS).