Applied Multivariate Methods for Data Analysts

Rating:
Summary: Light on understanding, heavy on computer output
Review: A great text to learn the mechanics of multivariate analysis. Unfortunately, a poor one to learn what exactly you're doing!

The author and Duxbury Press disappoint the reader with many instances of improper writing and editing. Occasionally, misused prepositions and descriptors couple with notation errors in formulae to leave the reader totally clueless. For a 1st edition, this may be expected.

The first chapter's lame attempt to introduce the field tries to accomplish more than is possible with a student just launching into multivariate analysis.

Many derivations do not appear; the text provides only a reference. The author could have included them in the appendix for the interested student. Instead, much of the appendix explains a huge social science dataset. However, a part of the appendix reviewing linear algebra proved useful for this student.

Most "figures" are computer printouts. The text frequently references parts of the printout far away, requiring the reader to flip pages too much. The chapter exercises also emphasize computer analysis, with nearly all requiring statistical computing packages such as SAS, SPSS, etc. Although these reinforce the mechanics of analysis, they neither challenge the student to understand the fundamentals of the techniques nor hone critical thinking skills. For example, some exercises are more amenable to short answers. Yet the text contains no "answers to exercises" section anywhere.

For the reader who needs some theory and a lot of "hands on" information, this text may work. But it misses the mark for aspiring statistical consultants by diluting statistical concepts with reams of computer output.

Rating:
Summary: I second the reviewer from phili 's review
Review: I hav e a stat background so like to see some theory, but this aside, I think most users will find it lacks in a decent level of backgound as to what is really going on. Works fine if you want a quick guide to how to get certain multivariate stat procedures done in SAS or soome other stat package

Rating:
Summary: An excellent book for practitioners of data analysis
Review: If you are in the business of analyzing data, and do not have a degree in multivariate statistics, this book provides you the tools you need to get started right away. The author seems to have deep hands-on experience with the methods he is presenting and gives insights on their use in practice, often through well chosen and developed examples, that a pure theoretician would simply lack. I recommend it wholeheartedly to everyone who earns a living by analyzing data.