Customer Reviews

Multivariate Data Analysis (5th Edition)

5 star:		(16)
4 star:		(2)
3 star:		(1)
2 star:		(1)
1 star:		(0)

 

 

I took multivariate in Ph.D. school. Our professor didn't have a book (just his notes), so I had no book to be loyal to when I got out. I was given this book once I graduated and consider it to be a fabulous reference. I can't speak to *learning* multivariate techniques out of it because that's not been my experience with this book.

I use it as a reference- to refresh myself on a technique, or to consult when I run into a problem- this book has yet to let me down and has been able to answer any question or solve any problem that I've had.

You see this book cited in academic behavioral research, but the book does a great job of explaining things in a managerial way as well.

Other of these reviews have criticized it for going on too long on an example or a technique- for that I PRAISE this book- I WANT that extra information. I'm reminded of that quote from "Amadeus"- "Too many notes." I WANT as many notes as I can get- that's what makes it so much more helpful.

If you are looking for a great reference book for multivariate techniques, look no further.

The Hair et al textbook has been one of my favorite recommendations to graduate students who are wrestling with their data analysis. I also use this book in a methodology course I give over here at Vrije University Amsterdam. The following points, in my opinion make this book the best choice for this purpose: 1. This book, contrary to many more statistically oriented texts, discusses analysis methods that are really used. Not the techniques that are necessarily most elegant, but the techniques that are most practical 2. The approach taken in this textbook is not formula-oriented. The authors simply discuss the prerequisites of a certain techniques and the interpretation of the outcomes. 3. The authors make a nice attempt to help the reader in choosing an appropriate technique. Although the treatment of the technique sometimes is too superficial, this is not a problem. A textbook cannot treat all details and good references are given in each chapter.

One the negative side, a number of caveats has to be mentioned however: 1. This most recent edition, unfortunately is worse than previous one. I tend to recommend older editions. The authors apparently have been trying to include a lot of material that users (i.e. instructors) of the book thought was missing. The structure of the book suffers from this added material. Another problem is that really high-level material gets mixed up with elementary statistics. A textbook simply cannot be everything for everyone. 2. The book is too big. Too much material has been added to this text. Furthermore, often too much time is spend discussing the examples. E.g., the illustration of discriminant and logit analysis alone takes over 40 pages.

Overall my rating of this book is not negative. If this edition of the book had been my first encounter with this text, I probably had been more enthousiastic. Now I tend to compare it with earlier editions, and must conclude the book has degraded. A next edition, in which a good editor goes through the text to enhance structure and get rid of superfluous material might easily solve the problems mentioned.

This is the best applied book on multivariate analysis I know. It clearly explains how to do statistical analyses and how to interpret the output. Clear examples throughout. Syntax supplied for each type of multivariate analysis in both SPSS and SAS, with LISREL notation for CFA. In addition to specific techniques (factor analysis, multiple regression, multiple discriminant analysis, MANOVA, conjoint analysis, canonical correlation, cluster analysis, multidimensional scaling, structural equation modeling/CFA) excellent sections on structuring data, cleaning data, and handling missing data. While mathematical sophistication always helps in stat, this book doesn't require it. No knowledge of matrix algebra needed to understand this book. Few if any formulas. Emphasis is on logic rather than math.

I've used this book (various editions) since graduate school -- more years ago than I care to admit. It is written in English rather than in "statistics" and gives a terrific conceptual overview of the research techniques. I keep it as a reference and use it regularly. If you want to understand multivariate stats without having to wade through a lot of symbols, this book is great. I also find it very useful when I'm trying to come up with a layman's explanation for management.

This book is an excellent source of information on multivariate analysis techniques. I especially like the flowcharts used for determining which analysis method to use as well as the flowcharts showing what steps to take for the analysis method chosen. If you have a good basic knowledge of statistics and a good head on your shoulders, you will have no problem understanding the methods presented.

A pretty good overview and a lot of in-depth material on Multivariate data analysis. Not quite a bed time read though.

I recommend this book as part of your analytical library.

If you liked this book, another good book on multivariate data analysis you may want to check out as well is Sharma, S.; Applied Multivariate Techniques, New York: John Wiley & Sons, Inc., 1996.

If you want something easier to read/more practical, and you prefer SPSS over SAS you may want to check out either `Discovering Statistics using SPSS for Windows' by Andy Field, or probably even better/simpler `SPSS Survival Manual' by Pallant.

Tom Anderson
Anderson Analytics, LLC
(...)

This is without question THE BEST introduction to Multivariate Statistics book currently available. It is designed for the user of the techniques, not someone who wants to examine the math underlying the techniques. I have created a collection of the various editions of this book and I have all of them going from the 1st edition to the current one. Personally I really likely the 2nd and 3rd editions but the current one is also very good. Whether you are interested in Exploratory Factor Analysis, Multiple Regression Analysis, Discriminant Analysis [I think that there should have been more on classification analysis in this section], Logistic regression, multivariate analysis of variance, conjoint analysis, cluster analysis, multidimensional scaling, Confirmatory Factor Analysis or Structural Equation Modeling, this book provides a good broad overview as to how to use and interpret the techniques. The key terms for each technique are defined clearly technique by technique. Having taught faculty how to teach multivariate statistics this is the book that I chose to use. It is important to remember that it is BROAD overview and if you are going to do serious analyses that you'd likely want to get additional books about the specific technique or techniques that you are going to use.

The more powerful and advance research techniques becomes, the more easy it is for an average analyst to go astray. This is what the text is aimed for: it helps build a very solid conceptual foundation for researchers when using multivariate analysis. Concepts are clearly delineated and criterias to using / NOT using a certain techique are also fully outlined (Chapter 1&2). Incidentally, this is not a step-by-step or so-called quick-fix manual that guides the readers what buttons to look for in the SPSS/SAS.

We used this book for our Stats 2 course in grad school, and although our professor was good enough to eclipse Hair, et al. (he had written his own Stats text), I am most pleased at how much mileage I have drawn from Hair, et al. in the years since I took the course. In my subsequent career, reviewers have often sent me back to Hair, et al. when they have questions about something I'm doing with a data analysis, so it has become an invaluable reference for that reason alone.

The chapters on structural equation modeling, MANOVA, and factor analysis are particularly useful and well-written. I recommend this book without reservations for graduate students and others who work with advanced statistics as part of their daily work.

Over the course of my undergrad, grad, and post grad, I have read a variety of statistics books. Without a doubt, Hair's Multivariate Data Analysis is THE BEST book of them all. Here is a brief outline of the awesome features of the book:

1. The book itself is very well organized - chapter order and the order within each chapter helps the reader in knowing what is coming next and provides a sense of direction. I think this is a very important feature for any book to have especially when the topics are complex and are discussed over 800 odd pages.

2. The HBAT data set that comes along with the book (or that is provided by the instructor of the course depending on the version of the book you purchase) is really a very good resource. All multivariate techniques in the book can be carried out using this data set. The data set is clearly explained at the end of the first chapter.

3. Tables of examples, the 'Rules of thumb" after each important concept discussion prove invaluable. This is akin to the managerial implication written at the end of lenghty academic articles. This is almost like saying - Here is the deal folks.....Much precise than the summary section, in bullet points, these rules of thumb acts as quick referece that captures the content of the discussion.

4. From chapter 4 onwards till the very end of the book, each chapter is divided into two halves - the first half is the concept dicsussion - in detail, with examples and in very simple and understandable language. The second half is the illustration of the discussed concept through a very elaborate example using the HBAT data set. This arrangement not only helps the reader in better understanding the complex concepts, but also allows the reader to get their hands dirty by actually working out.

5. Keywords at the begining of each chapter provides a list of all the 'jargon' that would be used in that chapter. This list provides a detail definition of each term. Many times while reading the chapter, you would come across a confusing term and in those times the keyword list can prove invaluable.

All in all, this is an invaluable book. If you are a taking stats and you have not read this book, you are missing something. In spite of all the above great things, the best feature of this book is the writing style. I have not come across a book that explains concepts is such easy to understand language but at the same time not over simplifying the subject matter.
My advanced stats became enjoyable because of this book. Really may GOD bless these authors for writing this book!!