An Introduction to Multivariate Statistical Analysis, 3rd ed. T.W. Anderson. Hoboken, NJ: John Wiley & Sons, 2003, 742 pp., $99.95, hardcover. ISBN 0-471-36091-0.

Abstract

This third edition is a wonderful textbook for graduate students studying statistics or professional statisticians interested in a thorough introduction to the mathematical theory underlying many common multivariate statistical methods. However, if you are a laboratory scientist lacking a strong mathematical background, this book is not for you.The author earned his PhD in mathematics at Princeton in 1945 and has been an influential figure in the statistical field for many decades. In addition to this book, he is also the author of The Statistical Analysis of Time Series. The first edition of An Introduction to Multivariate Statistical Analysis was derived from lecture notes used in a two-semester sequence of graduate courses given at Columbia University. Published in 1958, it was one of the early books on multivariate statistical analysis, along with M.G. Kendall’s monograph, A Course in Multivariate Analysis, published in 1957, and S.N. Roy’s Some Aspects of Multivariate Analysis, published the same year. It was intended for individuals interested in a “mathematically rigorous development” of the statistical methods. A background in univariate statistical theory and matrix algebra is a prerequisite.