2 comments about Statistical Methods for Spatial Data Analysis (Texts in Statistical Science Series).

1. This book was recommended to me by Amazon. It looks very attractive as I would like to learn more about this area.
   
   However, when I clicked on the sample pages online and read the first example, I realized there are typos already. First, the second covariance function should not have a conditional line, it is meant to illustrate the correlation or covariance between two observations in the same experimental unit, unconditionally. Second, I think the second covariance function is missing the tau_i which generates another variance component in the covariance function. This looks disappointing and I will not purchase the book now based on what I read.
   
   Author, any thought?



2. but accessible if you have the appropriate background. A great treatment of the underlying theory, without too much obfuscation. The authors include many interesting examples and make available SAS code on their website.

1 comments about Healthy People 2010 (softcover, combined volumes I and II) (USDHHS, Healthy People).

1. Like many documents from the Federal government, this book is not an easy read. The text is dense, complex, and full of jargon.
   
   However, it is critical that those in the area of health policy at any level (local, state, or Federal), cultural competence, or health literacy be aware of this outline of public health policy.

No comments about Event History Analysis with Stata.

5 comments about Multivariable Analysis: A Practical Guide for Clinicians.

1. I have to say I'm really impressed by this book. Its problem-based,user-friendly approach is a stand-out among all other statistic textbooks. The author did an excellent job in explaining the difficult concept of multivariable analysis using easy-to-read English and no jagons. Highly recommended for all.



2. Straightforward approach to the concepts of multivariate analysis (MA) in medicine. If you don't know anything about MA and wanna understand it in a easy and fast way, this is the book. Katz also helps you to perform your own analysis. However, it's important to point out that a solid basis on statistical methods are needed for choosing the best method in a particular situation. The Cox model is not the solution for all problems!!! No strong math background is needed. This is a book of concepts and not of techniques. At least I can guarantee you'll understand the multivariate analysis published in NEJM and will find out about how common multivariate analyses are inappropriately used in medical journals. A big plus that every physician must read!



3. This is one of the best books I have ever read about biostatistics. It takes you further -from the usually well known table statistics to the model statistics using a step by step approach. Without covering the overall field, it tries to fortell readers' questions and answers them in the most explicit way. It's a book to understand what is going on around journals and makes your steps easier once you decide to try your way to dive in your numbers.



4. I used this book for a class last year. I found the organization quite horrible. There are 3 main multivariate models that he discusses in the book: Multiple regression, logistic regression and Cox proportional hazards. He ignores ANCOVA, ANOVA and multi-level frequency table (chi-square) methods.

   Instead of writing about the 3 main multivariate models above individually he sporadically switches between all 3 models. The TOC is organized as a series of questions and answers. I would have preferred he wrote about each of the 3 topics separately.

   Although finding the information you want is difficult, the book is written superbly. He makes it easy to understand difficult concepts such as interactions, model building, collinearity and testing of assumptions. You don't need a math background to understand this book. Aside from the organization of the contents, I loved this book! I would recommend for clinicians who are interested in learning about how multivariate models are created. If you review a lot of manuscripts in medical literature, this is a must read.




5. From the start I felt intimidated by the title and even with the mitigating suggestion that this would be practical, a guide, and for the clinician, the challenge was there. Throughout our training, in textbooks and journals, and in many lectures, we have to analyze the information put in front of us. Our background in basic statistics is typically primitive, and certainly could use some help in understanding what we read. This very readable and even interesting 200 plus page book really fills that need. With an introduction as much seductive as informative, Dr. Katz makes his way through his early chapters discussing common uses, outcome variables, types of regression and other basic definitions. He uses the question and answer format which lends itself to learning the information in a building block manner. He constructs the book with numerous examples, handy tables and graphs and very helpful side boxes filled with pithy hints for making progress through the book. Each chapter has a built in summary and completes the questions asked. The index is complete and very pointed, making the reader able to use the book not only to learn the subject, but to refer back to previous reading. This book had an enthusiastic first outing, and certainly this second edition is worth the price for a good reference. Journal of the Kentucky Medical Association, Vol 104 (8), 2006, p.395

5 comments about Model Selection and Multi-Model Inference.

1. Burnham and Anderson have put together a scholarly account of the developments in model selection techniques from the information theoretic viewpoint. This is an important practical subject. As computer algorithms become more and more available for fitting models and data mining and exploratory analysis become more popular and used more by novices, problems with overfitting models will again raise their ugly heads. This has been an issue for statisticians for decades. But the problems and the art of model selection has not been commonly covered in elementary courses on statistics and regression. George Box puts proper emphasis on the iterative nature of model selection and the importance of applying the principle of parismony in many of his books. Classic texts on regression like Draper and Smith point out the pitfalls of goodness of ift measures like R-square and explain Mallows Cp and adjusted R-square. There are now also a few good books devoted to model selection including the book by McQuarrie and Tsai (that I recently reviewed for Amazon) and the Chapman and Hall monograph by A. J. Miller.

   Burnham and Anderson address all these issues and provide the best coverage to date on bootstrap and cross-validation approaches. They also are careful in their historical account and in putting together some coherence to the scattered literature. They are thorough in their references to the literature. Their theme is the information theoretic measures based on the Kullback-Liebler distance measure. The breakthrough in this theory came from Akaike in the 1970s and improvements and refinement came later. The authors provide the theory, but more importantly, they provide many real examples to illustrate the problems and show how the methods work.

   They also refer to the recent work in Bayesian methods. Chapter 1 is a great introduction that everyone should read. Being a fan of the bootstrap I was interested in their coverage of it in chapters 4, 5 and 6 (much of which is the authors' own work).

   Because the authors work in biological fields they cover survival models as well as the standard time series and regression models where most of the emphasis has been placed on model selection in the past.

   It is a great reference source and an important book for learning about model selection as part of the inferential process. The pictures of the famous contributors inserted throughout the book is also nice to see. We have Akaike, Boltzmann, Shibata, Kullback, and Liebler brought to life in photographs or sketches.




2. AIC is one of the widely known methods in model selection and inference.
   This book includes not only a basic use but also advanced issues of the information-theoretic approach.
   Using this book, you will learn the application of AIC soon!



3. I admire this book very much for its accessible treatment of AIC, but if were reduced in length by half, it would be twice as good. The authors cannot resist repeating themselves, usually several times, especially when giving advice of the "motherhood and apple pie" variety. Another annoying feature is that many references are given for philosophical points, yet sometimes when a useful result is given without proof, no reference is provided. For example, on page 12 an expression for maximized likelihood is given without a derivation or a reference. Inside this fat book there is a thin book crying to be let out.



4. Those interested in mark-recapture models definitely should have this extraordinary book.
   Very complete and easy to read



5. Burnham and Anderson have put together a scholarly account of the developments in model selection techniques from the information theoretic viewpoint. This is an important practical subject. As computer algorithms become more and more available for fitting models and data mining and exploratory analysis become more popular and used more by novices, problems with overfitting models will again raise their ugly heads. This has been an issue for statisticians for decades. But the problems and the art of model selection has not been commonly covered in elementary courses on statistics and regression. George Box puts proper emphasis on the iterative nature of model selection and the importance of applying the principle of parismony in many of his books. Classic texts on regression like Draper and Smith point out the pitfalls of goodness of ift measures like R-square and explain Mallows Cp and adjusted R-square. There are now also a few good books devoted to model selection including the book by McQuarrie and Tsai (that I recently reviewed for Amazon) and the Chapman and Hall monograph by A. J. Miller.
   Burnham and Anderson address all these issues and provide the best coverage to date on bootstrap and cross-validation approaches. They also are careful in their historical account and in putting together some coherence to the scattered literature. They are thorough in their references to the literature. Their theme is the information theoretic measures based on the Kullback-Liebler distance measure. The breakthrough in this theory came from Akaike in the 1970s and improvements and refinement came later. The authors provide the theory, but more importantly, they provide many real examples to illustrate the problems and show how the methods work.
   
   They also refer to the recent work in Bayesian methods. Chapter 1 is a great introduction that everyone should read. Being a fan of the bootstrap I was interested in their coverage of it in chapters 4, 5 and 6 (much of which is the authors' own work).
   
   Because the authors work in biological fields they cover survival models as well as the standard time series and regression models where most of the emphasis has been placed on model selection in the past.
   
   It is a great reference source and an important book for learning about model selection as part of the inferential process. The pictures of the famous contributors inserted throughout the book is also nice to see. We have Akaike, Boltzmann, Shibata, Kullback, and Liebler brought to life in photographs or sketches.

2 comments about Systematic Reviews in Health Care: Meta-Analysis in Context.

1. This book is written to be understood and used. Everything is easily accessible, even for the relatively innumerate, and presented in clear, easy English. The authors are to be congratulated on excellent pedagogical prose.
   
   The major downside was a skinny statistical section that lacked depth (what was present was VERY clear however). Since the majority of the book is low-level, big-picture explanations of the principles of meta-analysis, I think that an expanded statistical section would not detract. In fact, because the rest of the book is so well explained, I suspect that most readers would have a good appetite for implementing the ideas presented. Those not inclined to read more details on implementation could just skip this section. In particular, I wished the authors had discussed how to combine simple proportions and rates across studies. This is arguably the simplest form of meta-analysis and the one for which nearly no advice exists to guide practise! There are tons of ways to transform and combine rates/proportions and these should have been developed, compared and constrasted for different scenarios. Similarly, I wish that even more Bayesian methods would be presented.
   
   Lastly, the software section is old. New ideas using R, Revman, Bugs and other FREE software should be encouraged. Software that costs big dollars is probably not the best thing to present in an international book. Everyone should be able to reproduce the results at their home. Free software ensures this.



2. This classic book by the legendary experts in meta-analysis and systematic review, Mattias Egger, George Davey Smith, and Doug Altman, is considered the most pre-eminent and most definitive textbook in this area. No other books can compare! Very easy to read with clear explanations, this is the most trusty reference textbook to own when doing systematic reviews and meta-analyses.

3 comments about Bioinformatics and Computational Biology Solutions Using R and Bioconductor (Statistics for Biology and Health).

1. I purchased this book to learn specific details and look at applications for the functions present in bioconductor. I have had trouble applying some of the chapters to custom data because they are written for specific microarray/data formats. Overall, this book is a good value because it contains examples of how bioconductor can be used to aid in hypothesis testing, but I struggle to apply what I have read to the different types of data I have. The section on Statistical analysis for genomic experiments and the section on gaphs and networks should be the reason you purchase this book. They are very helpful and interesting. The case studies were not very helpful in my opinion.



2. If you're like me, you came upon this book because you decided to use R for analysis of microarray data, but you're mired in its gory and frustrating details.
   
   Yes, you need a reference book. But not this one, and certainly not this edition. Better documentation can be found elsewhere (dare I say online?).
   
   The code examples given are technically accurate and run as advertised, but they are of the "monkey see, monkey do" variety. They provide little intuition for how to use R for oneself, outside the covers of this text. For example, Chapter 23 discusses linear models for microarray data (using the "limma" package), and several code examples contain the parameter 'adjust = "fdr"'. The reader is never enlightened that this refers to a "false discovery rate" adjustment.
   
   In other cases, example code is simply missing. Chapter 21 covers the Rgraphviz graphing library, with a figure showing the three common graphical layouts -- but no example code for producing these graphs is given (I had to find it outside the book).
   
   For those trying to use R for computational biology, I recommend getting an overview of the R programming language first (Venables and Ripley's book "Modern Applied Statistics with S" is a great text), and only then wading into references such as this one, if at all.



3. I find this book is not so good for people without any gene or microarray experiment background. It didn't even give clear definition of the basic concepts.
   Another problem is that it's not well organized because every chapter is written by different authors who have different interest and preference and use slightly different terms for the same thing.

1 comments about Statistics In Kinesiology.

1. The book was sent within seven days, much before than I expected. I simply love this service...

5 comments about Biostatistics: A Foundation for Analysis in the Health Sciences (Wiley Series in Probability and Statistics).

1. I'm taking my first biostatistics class in medical school, but the text I am using in class (Rosner) has lost me. I subsequently borrowed Daniel's book from the library (only because it has 7th editions), and I am glad that I picked it! Daniel is a good writer. The book is well organized and laid out. Important concepts are emphased and explained with minimum mathematics involved. The well thought out examples are worth working through as well for clarification of the applications of important concepts. However, as a beginner in statistics, I was lost in the midst of mathematics on certain concepts (given that I have a relatively strong mathematics background) without really understanding the meaning of some very basic terms, like percentile, confidence intervals.
   
   What I do is to read another reference book that explains the very basic concepts in plain English first before reading this text. I am currently using Munro's Statistical Methods for Health Care Research. While both of them cover the same set of concepts, Daniel gives me the mathematical and more advanced explainations compare to Munro.



2. Daniel obviously knows his statistics; but, I wouldnt think that is too helpful for individuals reading or studying from his textbook.
   The reasons are numerous, and all these reasons would reduce anyone's chances of solely using this book, or even using it at all.
   The textbook is well organized, however Daniel's writing often is pedantic, repetitive (not in the helpful way) and ambiguous at best.
   The examples and solutions occassionaly have serious errors in them which affect the overall outcome of the test (A second consideration is that the book is in it's 8th edition!!! therefore such errors are unacceptable for a person such as myself).
   An example can be found on page 239 (example 7.3.2). The pooled variance, as calculated by Daniel is approximately off by 100 simply because he didnt give attention to dividing the numerator with the proper pooled D.F of the samples. The chapter ironically was on hypothesis tests, something extremely important to any line of empirically oriented statistics.
   
   In Chapter 8; which is probably the most important chapter in Bistatistics (ANOVAs) he does not mention the relationship between MSW and sample SD. Also, his usage of Summation in formulas often are unnecessarily overcomplicated. Such is not even seen in professional journals.
   
   I did like this textbook regardless of its many shortcomings, its not because I liked the author's style of writing. Its more or less the fact that my lecturer (I assume) used this book heavily in his lectures and so I used it as a supplementary text.
   
   I would suggest, Chap T. Le's Introductory Biostatistics. However he goes too much into nonparametric methods and proportions and doesnt cocentration (to the degree I wanted) on continous data.
   
   More robust and probably cost effect books are :Introductory Statistics for the Life Sciences by Samuels. But the Best book I have ever seen on the subject is "Introductory Biostatistics for the health sciences" By Chernick and Friis. The book is well priced and no portion of this book, I have seen as being useless.



3. This book is extreamly helpful for academic research. It can be somewhat more technical than most people would need.



4. I appreciate the quick and honest service I received. It was a very easy transaction.



5. Nice biostat book that works as a good reference for people with biostatistical interests. I bought it because I had a biostat course this semester. It's nice, I recommened it.

5 comments about Survival Analysis.

1. In this new edition, most of the errata are corrected and the texts are explained in a more detailed way.

   The formulae are correct and the examples are explained in a more direct and expressive way than that in the 1st edition.

   The most valuable one is its Theoretical Notes and Practical Notes. They show a lot of different points of views.

   A good-buy and must-read for those want to have an intense level in Survival Analysis. Suitable for elementary and intermediate candidates to read and study.

   Ian Lauder




2. This textbook is too heavy on mathematical theory, and as a result ends up being largely uninformative. It is also long-winded to the point of being interminable. In order to implement survival analysis techniques, the practicing statistician does not need to wade through endless proofs, derivations, and digressions into the specification of likelihood functions. There are many textbooks available that provide a more intuitive understanding of survival analysis techniques, in a much shorter space.



3. I am a computer scientist and using this book for my research to address a problem. This book is well written but of course target audience are people with solid background in probability theory and parameteric estimation (pattern recognition). Therefore please do not expect that author will teach you basic probability theory. Contents are more applied in nature therefore natural audience are staticians and researchers.



4. I used this book for a class in survival analysis (a graduate level biostats course) and I found it very useful. Much of the first several chapters are fairly quick relative to many graduate statistics texts and focuses on application with less emphasis on theory. Overall, I have no major qualms with the book. The author goes on a bit longer than necessary but I'd rather end up skimming text than be stuck deciphering terse material. This extra explanation also opens the book up to a wider audience.
   
   A solid understanding of basic statistics is necessary to get started in this book. To get more, 4+ semester-long statistics courses, at least one based in regression, would be ideal. A basic knowledge in mathematical analysis as it pertains to statistics (mainly dealing with convergence in law) will be beneficial to understanding some of the intricacies of the topics and answer many of the 'whys'.
   
   In conjunction with the course and the book, I worked problems in R with the 'survival' package, which I found very useful. (R is a free statistical program. A basic understanding of R would be necessary before trying to use the survival package -- I would recommend Dalgaard's book for an intro to R if this is of interest.) I have a good understanding of R and found the survival package documentation supplemented by rseek . org searches (when I got stuck) sufficient to figure out how to implement the survival functions in R.
   
   On the example setup and problems...
   at the end of each chapter, this book is a bit hit-or-miss. Some problems are good. Many are not. There is a lot of confusion created by some of the problems, which leads into the part of the book I take the most issue with. The authors refer to scattered examples in problems (take for example, referring to example 8.3 in problem 9.5). The thing is, Example 8.3 starts on page 251 and then it continues randomly throughout the remainder of the chapter until page 274 (I had to page through the chapter to find those page numbers). The examples in mid-to-late chapters can be very scatter-brained and some of the problems start to become this way as well. The authors seem to forget that keeping track of the 15-20 studies they use in this text is no small task and that they've spent a lot more time looking at them than others. Self-contained examples where I don't need to flip back to chapter 1 or some other example to read about the study would be really nice. The examples and problems could have been much more user-friendly to accelerate the learning process.



5. The authors present an intermediate level text on survival analysis introducing the concepts and techniques and providing many real examples. Covers all the standard methods including the Cox proportional hazard model (with stratification) and some methods not commonly covered including regression diagnostics and multivariate survival methods (including fraility models).