The goal of Norman H. Anderson's new book is to help students develop skills of scientific inference. To accomplish this he organized the book around the "Experimental Pyramid"--six levels that represent a hierarchy of considerations in empirical investigation--conceptual framework, phenomena, behavior, measurement, design, and statistical inference. To facilitate conceptual and empirical understanding, Anderson de-emphasizes computational formulas and null hypothesis testing. Other features include:
*emphasis on visual inspection as a basic skill in experimental analysis to help students develop an intuitive appreciation of data patterns;
*exercises that emphasize development of conceptual and empirical application of methods of design and analysis and de-emphasize formulas and calculations; and
*heavier emphasis on confidence intervals than significance tests.
The book is intended for use in graduate-level experimental design/research methods or statistics courses in psychology, education, and other applied social sciences, as well as a professional resource for active researchers. The first 12 chapters present the core concepts graduate students must understand. The next nine chapters serve as a reference handbook by focusing on specialized topics with a minimum of technicalities.
Table of Contents
Contents: Scientific Inference. Statistical Inference. Elements of ANOVA-I. Elements of ANOVA-II. Factorial Design. Repeated Measures Design. Understanding Interactions. Confounding. Regression and Correlation. Frequency Data and Chi-Square. Single Subject Design. Nonnormal Data and Unequal Variance. Analysis of Covariance. Design Topics I. Design Topics II. Multiple Regression. Multiple Comparisons. Sundry Topics. Foundations of Statistics. Math Models for Process Analysis. Toward Unified Theory. Principles & Tactics for Writing. Lifelong Learning. Basic Statistical Concepts.
"...Anderson's latest book is truly remarkable. It is a graduate level text with the first 12 chapters covering core principles and the remaining 9 chapters serving as an excellent reference on numerous statistical methods and design issues....The book captivates your attention as topics are addressed from a perspective you've never entertained when studying statistics and experimental design."
—Air Force Research Laboratory-Science and Technology
"Unlike most statistics texts, this one strives to be both educational and polemical--the author wants most graduate programs teaching statistics to be more empirically based (no new claim) so that future psychologists will acquire sounder research judgment. Novelty lies, however, in the integrated approach Anderson offers for developing such judgment....This book is clearly the culmination of a career devoted to rigorous reliance on the experimental method, meticulous and thoughtful planning of research, and a delight in solving cognitive-behavioral puzzles."