Mathematical Statistics and Data Analysis

Rating:
Summary: The best study material for statisticians
Review: Rice's book is well written and statistical ideas are communicated nicely to the reader. You need to have a good mathematical background to use this book because many theorems are proved intuitively in this book. However, Rice failed to communicate ideas behind the Neyman Pearson Lemma in chapter 9. Do not be surprised if it is difficult to understand that chapter. It is also important that your basic statistics knowledge is good before attempting to use this book. For sure, your problem solving skills will improve soon as you start solving problems from this book. On average, the book is very good. I have used it to refresh my statistics knowledge when doing my MSc Engineering Mathematics degree.

Rating:
Summary: Lacks rigor, but eminently practical
Review: Rice's _Math. Stat. & Data Analysis_ is ideal for students in the social sciences (particularly experimental psychology). The text provides enough relevant theory so that the reader can practice statistics selectively and informedly, rather than blindly substituting values into formulas. However, those with a strong theoretical interest are ill-served; the "proofs" are numerous but not rigorous, and some interesting results are hand-wavingly assumed to be true.

Rating:
Summary: Is there an ideal text that non-statisticians will love?
Review: 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.

Rating:
Summary: Complex and Vague
Review: The book covers very complex concepts (includes some very difficult applications), whereas lacked the explanations of some very basic concepts (how to use statistical tables etc.) The book is definitely not for beginners (prob. theory and statistics). The disk that was included did not work; probaby requires some special software. Good aspect of the book was that it included lots of exercises. Recommended for someone that has already studies statistic for at least a semester. Beginners; forget it!

Rating:
Summary: Not Recommended
Review: The book is badly written. Very Very bad design. Although it is full of content and not you could find loads of information about
theorytical and practical statistics, the language is often confusing and important data are given in the middle of redundant paragraphs. I wish the book had sections with key information and the examples where better explained.
A person with limited math-stats background will be lost.

Rating:
Summary: poorly written and difficult to understand
Review: The chapters are poorly written and difficult to read. The book does a good job of taking simple concepts and making them practically incomprehensible. Many of the examples refer to and build on previous examples, so if you get lost on one example it becomes even more difficult to understand anything that comes after it. And on top of all this, the problems at the end of the chapter often seem to have little to do with the material that is "covered" in that chapter. Overall, a bad book that made my prob/stats class much more painful than it should have been.

Rating:
Summary: nice introduction with some important advanced topics
Review: 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.

Rating:
Summary: Great Reference
Review: This book is not for the beginner. It is not for someone who doesn't already have a good deal of elementary probability as part of his 'blood and bones.' It is not for those without a functional knowledge of multivariate integration and the transformations involved therein: just as for any "mathematical statistics" textbook, mathematical competence is critical in deriving utitity. This is not a probability textbook in the vein of Stephen Ross (whose text would be best to consult first if one is lacking in such a background.) In short, if one lacks facility with calculus, one should look elsewhere first.

But then return. Return to this wonderfully complete and rigourous txt that offers challenging end of chapter exercises and insures that if you look something up, you will find it. The contents is vast, and each time you open it, you're likely to walk away with something new or something appreciated fully for the first time.

Misses the 5th star because of its list of errata, and because newer editions haven't been forthcoming. It could use one more revision.

Rating:
Summary: Don't believe the bad reviews of this book
Review: 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.

Rating:
Summary: Simply one of the worst textbooks in statistics
Review: This book is the worst math / statistics textbook that I have ever encountered. The explanation for most concepts is either inadequate or incomprehensible. The worst fact is that Rice has a tendency to introduce new variables in examples and proofs without ever defining the variables. Rice's notation also differs from standard statistical notation for many topics. Overall, there are really very little redeeming features about this book except nice tables and several good problems. I strongly do not recommend purchasing this book, as it makes a worthless reference.