[Home] [Table of Contents] [External Reviews] [My Own Review] [The CALSTAT Suite of Programs] [Data Entry Window] [What is Functional Measurement?] [ANOVA FAQ] [Report problems with CALSTAT or text] [Purchase Information]

Analysis of Variance Textbook

Analysis of Variance and Functional Measurement: A Practical Guide

Teaching statistics has been a major focus of my professional life throughout my career. When I took a class in Analysis of Variance (ANOVA) from a fabled instructor, Norman H. Anderson, I loved the course but hated the textbook. The material appealed to my sense of order; I saw the elegant logic underlying factorial designs. But the text was a 500+ page tome filled with arcane formulas. There seemed to be no structure in the book, just one chapter after another filled with increasingly more complicated expressions. So I ignored the book and listened to the instructor very intently.

Over the years I taught the course, I developed my own course notes and eschewed a textbook.  Eventually those notes became what is, in my humble opinion, the best ANOVA text in the world. Students say it's a joy to read (or at least as joyful as a statistics book can be). The examples are worked out in detail, and the exercises are realistic in theme (albeit with small data sets - I'm not teaching how to type numbers) and sometimes even amusingly written.

What makes this text a "Practical Guide", and what distinguishes it from most other statistics texts, is that the presentation is intended not to impress my peers, but instead to make things clear to students to whom these concepts are new. The algebra is kept to a minimum; I view ANOVA logic in terms of patterns rather than formulas. It is written for people who want to know what to do and to understand why the procedures make sense. It is pitched toward psychologists and other behavioral scientists, not toward statisticians.

When Visual Basic became available, I wrote my own suite of WINDOWS programs for ANOVA. I didn't like the commercial programs; they seemed to require a separate course of their own. The most popular program among psychologists, SPSS, is perhaps the worst offender. I thought ANOVA software should be intuitive, easy to use. And the program should not open with a blank screen! 

My suite is called the CALSTAT series, named after the university where I have done my teaching. Students universally produce correct tables within five minutes of opening the ONEWAY program. I can devote my class time to teaching statistics rather than to teaching how to operate the program. Here's a screen shot of the data entry window for the factorial ANOVA program (all of the programs use the same interface).