Book Reviews:Statistical Analysis in Psychology and Education (sixth edition)

I was glad to review the sixth edition of Statistical Analysis in Psychology and Education by Ferguson and Takane. I have used previous editions (Ferguson, 1976, 1981) as introductory and intermediate level statistical texts. In a way, reading the sixth edition was like having a conversation with an old friend. What I remember about previous editions of this book is true of this one. Statistical vocabulary, notation, and concepts are explained in an intuitive, nontechnical manner. The ability of the authors to communicate complex ideas to students is an asset for any textbook. Other pleasures retained in the sixth edition are sections on the origins of the statistical ideas in the book. Whether it is a quotation from Darwin about natural selection in Chapter 1 or a summary of Spearman's theory of two factors in Chapter 28, the book puts these ideas in a historical perspective. The book has 28 chapters and 13 appendexes. The chapters are grouped into four parts: Basic Statistics, with 13 chapters; The Design of Experiments, with 7 chapters; Nonparametric Statistics, with 2 chapters; and Psychological Test and Multivariate Statistics, with 6 chapters. Part 1 includes material typically covered by a one-semester introductory course. Chapter 1, the introduction, gives an overview of data reduction, inference and estimation, types of variables, scales of measurement, and exploratory versus confirmatory data analysis. Chapter 2 covers the vocabulary of frequency distributions and the conventions about their tabular and graphical display. Displays such as the stem and leaf or the box and whiskers plots are missing from this chapter. Surely these graphical forms are as important as histograms or frequency polygons in modern statistics. Chapter 3 is a concise description of the summation rules. The notational conventions for frequency distributions are also in this chapter. Chapters 4 and 5 cover the measures of central tendency, variation, skewness, and kurtosis. A noteworthy section is on the mean as the point that minimizes the sum of squared deviations. This introduces least squares early in the text. Moments about the mean, skewness, and kurtosis are described. The formula for the kurtosis measure is incorrectly given as the ratio of the first moment to the squared second moment, minus three. The fourth moment should be in the numerator of the ratio. Chapters 6 and 7 describe probability concepts, the binomial distribution, and the normal curve. Chapter 6 is clearly written. Bayes' theorem