Statistical Methods for the Social and Behavioral Sciences

Abstract

Statistical Methods for the Social and Behavioral Sciences consists of 50 chapters organized into 9 parts. In Parts III-VII, statistical procedures are presented by distribution: methods based on exact distributions (Part III), normal distribution (Part IV), chi-square distribution (Part V), t distribu- tion (Part VI), and F distribution (Part VII). Exercises are included at the end of each chapter; answers selected problems are presented at the back of the book. A list of additional readings, arranged by parts, appears near the end of the book. Definitions are indicated in the index by boldface numbers; procedures are marked by italic numbers. The book was for the second statistics course in psychology, education, and other social science disciplines.., .and is well-suited both one- and two-semester courses.... The book progresses through an exten- sive of two-factor designs and a preview of higher designs analy- sis of covariance, regression analysis, and log-linear analysis (book jacket). Multiple sample tests are addressed prior of variance. Para- metric and nonparametric methods are presented on an equal basis, and concepts are developed as they are needed. The purpose of and assump- tions underlying each statistical technique are emphasized. A step-by-step how to summary of statistical procedures is given in boxes in each chap- ter. Examples stress application and interpretation. Simple algebraic developments are included in the body of each chapter; more advanced developments appear in a technical appendix at the end of the chapter. The book has several strengths. It is well written and easy understand. Mathematical skill demand is minimal; a one-semester high school or un- dergraduate college algebra course will enable the reader work through the book. Concepts are presented simply and are built on throughout suc- cessive chapters. Many presentations are intuitively appealing and elegant in their simplicity. The text is remarkably free of spelling and formula errors. (The odds-ratio formula on p. 711 is incorrect; see Kleinbaum, Kupper, & Morgenstern, 1982; Knoke & Burke, 1980.) Boxes are used present an easy-to-understand, step-by-step summary of statistical proce- dures. Parallel treatment of parametric and nonparametric methods is well