Customer Reviews

Contemporary Psychometrics: A Festschrift for Roderick P. Mcdonald (Multivariate Applications Books) (Multivariate Applications Series)

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

 

 

The book is an absolutely wonderful tribute to Rod McDonald. Like his work, it covers all areas of psychometrics from classical test theory to item response theory to covariance structure modeling. The topics range from those of historical interest such as rotational indeterminancy to the most state of the art advances in the field such as modeling time series processes. Other chapters integrate topics into cohesive frameworks, such as the chapter by Marsh, Hau, and Grayson on goodness of fit, the chapter by Harlow which focuses on the themes underlying multivariate statistics, and the chapter by Maydeu-Olivares which provides a unified framework for item response theory and covariance structure modeling. I highly recommend the book to those in the field of psychometrics; it is an indispensable resource for the field.

I was one of Rod's students in his last semester at Illinois. He was a wonderful teacher for me and I feel deeply honored to have been able to learn from him. Anyone who wants to understand factor analysis could do no better than getting a copy of Factor Analysis and Related Methods. It's 20 years old now, but nonetheless I hold it up to any other book on the topic. I think it is fair to say that the community is looking forward to Rod's long-awaited book on SEM.

Albert Maydeu-Olivares (another Illinios grad) and Jack McArdle did a very good job picking essays from the conference in Rod's honor held a few years back. These essays cover a lot of territory---like Rod's career. Some topics were more interesting to me than others, of course, but I felt like I learned something from them all. In particular, the chapters by Terry Ackerman, Dan Bolt, Albert, and Wolfgang Hartmann were memorable. (Factor indeterminacy is, still, about as interesting as a root canal, however.)

As to the one negative review: If the back cover lead you to believe this book was an introductory work, that's too bad. Like most other edited volumes, it's really a book for people who are in a bit. It would be a nice thing for say, a third year in a psychometrics program, but a first year would be totally lost. Hopefully you got your hands on a more useful book.

This is indeed a tour-de-force through contemporary Psychometrics. On the Wow! side.
The book contains many true gems, like the chapters on resampling methods in structural equation models, time series modeling through dynamic factor analysis, or on the RAM model for structural equation modeling. I also loved the chapters on limited vs. full estimation of IRT models, and on Bayesian methods for factor analysis.

It is a must read for anyone interested in finding out what Psychometrics is really about.

This book is a nearly worthless collection of essays. I suspect the two folks who gave it five stars contributed it to it. Notice that they have not bothered to put their real names or to review anything else.

The problems I have with the book include the following:

1) It's not advertised as a collection of academic essays, but that is all it is.

2) It is not suitable for reference.

3) It is not an accessible introduction to psychometrics for someone new to the field.

4) It is not comprehensive.

On the plus side:
The essays are written by psychometricians for psychometricains, most of whom hold advanced degrees in this very specialized subjects.

Overall: This thing is a complete waste of money for most of us. I wish I would have sent it back as soon as I received it.