Customer Reviews

Applied Regression Analysis and Generalized Linear Models

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

 

 

Now that I have had a few more classes in the subject area, I feel a bit more confident that this book should have an average rating, rather than higher. The explanations of the book are not bad, if you already have a thorough understanding of the topic and are using this is a reference. It does provide a quick overview of most of the major topics in the field and includes a full chapter on the treatment of statistical analysis using matrices and graphical vector visuals.

However, the organization is poor. Linear algebra, matrices and vectors should be introduced in the more accurate place of chapter 3,4 versus far later. Further, as a teaching tool, this book lacks practice problems to help the student through the learning process relative to other pieces that I've used. Further, each topic is addressed in the brief, which is good if you know the topic, but bad if it's the first time you're really looking at the work. The examples used are a bit discipline specific, that while not obscure, would make it somewhat difficult for newbies to the field to really obtain the type of practice and deep understanding that is required to go on to the next topic with confidence.

Higher rated texts should split up the topics into multiple books or provide a greater number of examples and problem sets for students to build their skill set. I have seen some that include an accompanying CD of data and practice examples, that can be of great assistance to students struggling to learn this discipline.

John Fox is a brilliant mathematician, but a poor author. The most frustrating aspect of his book is the failure to specify each component of each equation, especially when they build upon each other. After completing the associated course I am still uncertain what the difference between Xi and Xj is. He introduces "M" in chapter 6 to explain the properties of the least squares estimators, but does not explain that M means matrix until chapter 9--a chapter which is intended for the mathematically initiated (instead of forewarning students, chapters 9 and 10 simply should have been left out of this text altogether). There are also far too many instances of circular referencings to detail in this review. Simply put, if you are a professor considering using this text, be certain that your expected students have completed two semesters of calculus at the undergraduate level and two statistics courses at the graduate level. This book is perhaps best suited for special topic courses for students seeking a more profound understanding of general linear models. In conclusion, there are far more accessible texts available that address the major concepts of regression and mulivariate analysis.