Factor Analysis: Classic Edition

Psychology Press & Routledge Classic Editions

by Richard L. Gorsuch

Comprehensive and comprehensible, this classic text covers the basic and advanced topics essential for using factor analysis as a scientific tool in psychology, education, sociology, and related areas. Emphasizing the usefulness of the techniques, it presents sufficient mathematical background for understanding and applying its use. This includes the theory as well as the empirical evaluations. The overall goal is to show readers how to use factor analysis in their substantive research by highlighting when the differences in mathematical procedures have a major impact on the substantive conclusions, when the differences are not relevant, and when factor analysis might not be the best procedure to use.

Although the original version was written years ago, the book maintains its relevance today by providing readers with a thorough understanding of the basic mathematical models so they can easily apply these models to their own research. Readers are presented with a very complete picture of the "inner workings" of these methods. The new Introduction highlights the remarkably few changes that the author would make if he were writing the book today.

An ideal text for courses on factor analysis or as a supplement for multivariate analysis, structural equation modeling, or advanced quantitative techniques taught in psychology, education, and other social and behavioral sciences, researchers who use these techniques also appreciate this book’s thorough review of the basic models. Prerequisites include a graduate level course on statistics and a basic understanding of algebra. Sections with an asterisk can be skipped entirely if preferred.

Genres: Mathematics

455 pages, Kindle Edition

First published January 1, 1974

Reviews

[Terran M · Rating 5 out of 5]

This book holds up very well after 35 years, and is readable and useful today. Most data scientists have encountered the idea of principal component analysis, so I'll explain factor analysis relative to it. Factor analysis considerably extends PCA, by allowing factors which are correlated with each other (and then higher-order factors based on these correlated factors). It also allows for the possbility of explaining only communal variance (that captured by at least two variables) instead of all variance. The third extension is that in factor analysis, the factors or components may be post-rotated after they are extracted and truncated, to re-align them better with the original variables. This book explains all of these techniques and gives considerable empirical wisdom on which should be used when.

One caveat I will offer is that this book is useful only for exploratory factor analysis. Although there is a chapter on confirmatory factor analysis (testing a clearly defined hypothesis about the structure), it is too brief and does not sufficiently explain how the analysis is actually conducted.

[Eleanor Carson · Rating 5 out of 5]

This classic book is hard to reach now so being able to access a reissue of it is really helpful. I got this book to puzzle my way through explaining the theoretical underpinnings for decisions that turned out to work statistically. At the time I did the statistics I was following recommendations in a journal, but it wasn`t until I read Gorsuch`s book that I could understand why the method I followed worked so well. Some sections of this book are very technical (i.e. math-dense) and difficult to understand, but the theoretical discussions are enlightening.