This book integrates methods and data based on interpretations relevant to multivariate analysis. It uses an applications approach and is graphically oriented and strong on examples.

In recent years, innovations in computer technology and statistical methodologies have dramatically altered the landscape of multivariate data analysis. This new edition of Methods for Statistical Data Analysis of Multivariate Observations explores current multivariate concepts and techniques while retaining the same practical focus of its predecessor. It integrates methods and data-based interpretations relevant to multivariate analysis in a way that addresses real-world problems arising in many areas of interest.

Greatly revised and updated, this Second Edition provides helpful examples, graphical orientation, numerous illustrations, and an appendix detailing statistical software, including the S (or Splus) and SAS systems. It also offers

An expanded chapter on cluster analysis that covers advances in pattern recognition
New sections on inputs to clustering algorithms and aids for interpreting the results of cluster analysis
An exploration of some new techniques of summarization and exposure
New graphical methods for assessing the separations among the eigenvalues of a correlation matrix and for comparing sets of eigenvectors
Knowledge gained from advances in robust estimation and distributional models that are slightly broader than the multivariate normal