Book Review: Mendoza: Lomax, R. G. (2007). Statistical Concepts: A Second Course (3rd ed.). Mahwah, NJ: Lawrence Erlbaum

This is an introductory statistics book appropriate for an undergraduate or advanced undergraduate course in applied statistics. It could be used at the graduate level, but some of the material in the chapters would have to be supplemented. The book is very readable and requires little mathematical background. The author discusses the implementation of the statistical techniques in SPSS by giving examples and by showing the appropriate steps necessary to implement the techniques. For the most part, the book is formula-free emphasizing statistical concepts and not computations. The author covers the typical topics cover in an applied statistics course: analysis of variance (ANOVA), multiple comparisons, analysis of covariance, and multiple regression. The ANOVA is covered in detail and includes the discussion of designs with fixed, random, and nested factors, including some of the repeated measures designs. The coverage of these designs is very traditional, but the description of the designs is easy to understand. The discussion of assumptions and their violations is very comprehensive and up-to-date, including many references to relevant sources. Effect size calculations are discussed within the one and two-way ANOVAS but not beyond. Beyond the two-way ANOVA, we are shown how to obtain the effect sizes using SPSS. No formulas are given past the fixed two-way ANOVA. The coverage of MCP (multiple comparison procedures) is excellent. The concepts are clearly explained and the procedures are easy to understand. The author goes out of his way to tell us when each procedure is appropriate and when it is not, clearly differentiating between planned and post hoc procedures. The post hoc procedures are then categorized in terms of probability of type one error and power. After the MCP are introduced in chapter 2, they are discussed throughout the book as they apply to different designs. This book is clearly strong in its coverage of MCP and students as well as researchers will benefit from reading this chapter and the sections included in other chapters. Although the ANOVA designs are generally covered very well, there are some minor weaknesses in the coverage of random and mixed designs. For example, the justification of the F tests in the random models is done without the concept of expected mean squares. Although I understand that the author is trying to keep mathematical complexity to a minimum, the justification for the new denominator in the F ratios is hard to follow without the expected values. We also do not find much discussion of power in these complex designs. The section on randomized mixed effects is not very comprehensive, failing to include any reference to multilevel mixed linear models and their computer applications as in SAS Proc Mixed. The multivariate ANOVA is not fully discussed, although it is mentioned as a possible alternative to these univariate ANOVA models especially in the case of repeated measures. The chapter on randomized block designs is easy to read and very informative. The author discusses in detail different ways of forming blocks as well as the advantages and Organizational Research Methods 13(4) 838-839 a The Author(s) 2010 Reprints and permission: sagepub.com/journalsPermissions.nav http://orm.sagepub.com