Synopses & Reviews
This books presents an easy to read and wide-ranging introduction to techniques in multivariate analysis. It covers all the traditional topics of multivariate analysis including multidimensional contingency tables, logistic regression, cluster analysis, multidimensional scaling, and correspondence analysis. It is the companion volume to Volume I: Regression and Experimental Design published in 1991. The emphasis on the practicalities of the subject, and the author has included numerous analyses of real data sets drawn from a wide range of business, social sciences, and biological sciences settings. There are also many exercises which are designed to extend the analyses of the data sets including the use of statistical computing packages, and to cover further theoretical results relevant to the book. As a result, any student whose work uses these techniques will find this to be an excellent introduction to the subject.
"On the whole this volume on applied multivariate data analysis is a comprehensive treatise which will support students and teachers to a full extent in their coursework and researchers will find an easy ready-made material for the analysis of their multivariate data to arrive at correct conclusions. This is a masterpiece text." (Zentralblatt fuer Mathematik)
A Second Course in Statistics The past decade has seen a tremendous increase in the use of statistical data analysis and in the availability of both computers and statistical software. Business and government professionals, as well as academic researchers, are now regularly employing techniques that go far beyond the standard two-semester, introductory course in statistics. Even though for this group of users shorl courses in various specialized topics are often available, there is a need to improve the statistics training of future users of statistics while they are still at colleges and universities. In addition, there is a need for a survey reference text for the many practitioners who cannot obtain specialized courses. With the exception of the statistics major, most university students do not have sufficient time in their programs to enroll in a variety of specialized one-semester courses, such as data analysis, linear models, experimental de sign, multivariate methods, contingency tables, logistic regression, and so on. There is a need for a second survey course that covers a wide variety of these techniques in an integrated fashion. It is also important that this sec ond course combine an overview of theory with an opportunity to practice, including the use of statistical software and the interpretation of results obtained from real data."