This book presents in compact form a framework based in probability theory and the general linear model family for students and researchers using regression and analysis of variance methods. Special emphasis is placed on problems of properly using statistical computer programs. The relation between regression and analysis of variance is developed by means of the theory of linear contrasts for the benefit of students and users not versed in matrix algebra. Much attention is given to choosing proper error estimates, calculating proper estimates of standard errors in a variety of designs, and dealing with the problems of unbalanced designs. Having taught research design and quantitative methods in psychology for many years, Estes has developed ways of simplifying the presentation of concepts and derivations so as to make the substance of important statistical results available to students and investigators who lack much mathematical background and/or much taste for doing derivations.
Designed to supplement standard texts used in graduate courses in intermediate and advanced statistics, research methods, and experimental design for psychologists or other behavioral scientists, this text also has something to offer experienced investigators: material on model testing and related topics not covered in textbooks or other readily available sources.
"Estes [has] succeeded in making previously tedious material available to mathematically unskilled students."
"There's considerable good sense to be found in this book. The book is effective in its treatment of the one-way layout and noteworthy in its discussion of the two-way layout….Estes is particularly effective in ramming home the points that neither a statistical model nor a scientific model is a perfect representation of reality, and that a statistical model is only relevant to the extent that it illuminates a scientific question….Teachers of 'psych stat' could benefit from this book. A thoughtful scholar distilled considerable insight into its pages."
—Journal of the American Statistical Association
Contents: Introduction. Statistics, Probability, and Decision. Basic Concepts. Contrasts on Means. Testing a Statistical Hypothesis. Simple Analysis of Variance. Regression and ANOVA in the Linear Model Framework. Two-Way Factorial Designs. Repeated-Measures Designs. Unbalanced Designs and Nonorthogonality.