Mathematical Statistics and Data Analysis [Hardcover]

John A. Rice (Author)

Book Description

ISBN-10: 0534209343 | ISBN-13: 978-0534209346 | Publication Date: June 1, 1994 | Edition: 2

This is the first text in a generation to re-examine the purpose of the mathematical statistics course. The book's approach interweaves traditional topics with data analysis and reflects the use of the computer with close ties to the practice of statistics. The author stresses analysis of data, examines real problems with real data, and motivates the theory. The book's descriptive statistics, graphical displays, and realistic applications stand in strong contrast to traditional texts which are set in abstract settings.

Editorial Reviews

About the Author

Ph.D. University of California Berkeley

Product Details

• Hardcover: 672 pages
• Publisher: Duxbury Press; 2 edition (June 1, 1994)
• Language: English
• ISBN-10: 0534209343
• ISBN-13: 978-0534209346
• Product Dimensions: 9.3 x 7.5 x 1.3 inches
• Shipping Weight: 2.7 pounds
• Average Customer Review: 2.5 out of 5 stars  See all reviews (32 customer reviews)
• Amazon Best Sellers Rank: #322,469 in Books (See Top 100 in Books)

Customer Reviews

Most Helpful Customer Reviews

58 of 60 people found the following review helpful:

4.0 out of 5 stars Is there an ideal text that non-statisticians will love?, April 2, 2001

By 

Robert E. Froese (Houghton, MI USA) - See all my reviews
(REAL NAME)   

This review is from: Mathematical Statistics and Data Analysis (Hardcover)

Teaching statistics is a tough business because it is quantitative, rigourous, and often abstract. Most importantly teaching statistics is tough because the majority of students most professors face take statistics because the need to, not because they want to. To make matters worse, they face instructors who not only grasp the theory, but enjoy it, and who are all the while empowered to deliver no more than little snippets of higher level stuff their students can apply.

Rice tries to bridge the gap between theory and application, delivering enough theory that the student understands the logical foundation of the applied aspects they may have already discovered in previous courses. In my mind, this is the central theme of Rice's text - avoiding unnecessary and often pedantic details better left to graduate majors in statistics while filling in the background material that often left students of statistics uncertain about the amount of confidence to place in their analyses. Rice's text is not for those who fear rigour and logic. His introductions to new concepts are compact, impersonal, and often followed by terse propositions, definitions and laws that build logically as the text progresses. He includes numerous examples that are similarly terse; however, he never failed my litmus test for logical works, which is a demonstrable linkage between each example and some proposition, law or definition previously introduced.

The text commences with the most basic review of probability, progressing quickly to random variables, distributions, expected values and important derived distributions like the t, F and Chi-square. Students will discover how the tests they applied in the past are related to theory. This theme culminates in the section on Survey Sampling, in which sampling estimators and their assumptions are derived.

Rice has weaknesses that deserve mention. Some of the problems are tough, and Rice's impersonal approach emphasizes concepts over technique. I spent many hours reading and re-reading sections in the text before a useful approach to a problem came to me. Sections on least squares and ANOVA are the least useful; they are too compact to achieve the goal of bridging theory and application. This material is much better covered elsewhere. The decision theory and Baesian inference section suffers similarly, but given how little exposure most stats students get to this material is nevertheless useful.

If you're interested in learning the rigourous application of statistics but not theory, then Rice isn't for you. No matter what, you mustn't be afraid of challenges; Rice is impersonal and compact and won't make any excuses for you. If you want to understand the assumptions and limitations of the applied statistics you've already been practicing, however, I recommend Rice enthusiastically. He won't explain the assumptions, but he will arm you with the knowledge to do it yourself.

43 of 47 people found the following review helpful:

4.0 out of 5 stars excellent text, February 7, 2008

By 

Michael R. Chernick "statman31147" (Holland PA) - See all my reviews
(REAL NAME)   

This review is from: Mathematical Statistics and Data Analysis (Hardcover)

This book got very mixed reviews from 1 star to 5. I am in agreement with Froese's review and give it 4 stars. Rice is trying to write a book for statistics students who are not mathematics or statistics majors without shortchanging them on the advanced topics and the theory. This can be difficult and often alienates both the beginners and those interested in advanced methods. I have tried to stay along that fine line with my texts also. So I appreciate the difficulties. As an author of a book on bootstrap methods, I also appreciate the way Rice has integrated that subject into this text.

52 of 68 people found the following review helpful:

5.0 out of 5 stars Don't believe the bad reviews of this book, March 3, 2003

By 

Y "yyy" (United States) - See all my reviews

This review is from: Mathematical Statistics and Data Analysis (Hardcover)

This book is so far the best mathematical statistics and data analysis textbook I've ever read for an undergraduate intermediate level statistics course. The topics are well chosen and the book is well written. The previous bad reviews of the book at Amazon.com are from people with absolutely no knowledge of statistics and trying to find some short-cut to "prepare for a exam" or whatever. So if you are a serious reader and with intermediate level statistics understanding, go for the book. It is not only good to be used a textbook, but also excellent for reference purpose.