Customer Reviews

Introductory Statistics with R

5 star:		(10)
4 star:		(4)
3 star:		(4)
2 star:		(1)
1 star:		(0)

 

 

Introductory Statistics with R is an important book for a rapidly developing field. R is an extremely powerful statistical computing environment which suffers from the same problem as almost every other free software project -- a lack of quality documentation. Dalgaard fills a major gap with this book, that is, a guide to using R for many standard statistical problems.

For some time now, users have had to make do with S-PLUS books which contained some overlap with R. Now R users have a book they can call their own. After briefly discussing the R system and the language basics, Dalgaard goes through what might be covered in an advanced undergraduate data analysis course. Throughout the book, code examples and output are carefully interspersed so that the reader doesn't go too long without having a concrete example.

Dalgaard leaves out some advanced topics such as time series, spatial statistics, etc. (some of which are nicely covered in Modern Applied Statistics with S by Venables and Ripley) but that is probably for the best. The book is not bloated, nicely priced and I would recommend it to any advanced undergrad or first year grad student wanting to learn how to do statistical analysis in R.

Despite the web, there are learning curves sufficiently steep that a well-organized book is the most effective introduction. However, too many of these introductions, particularly in programming and/or statistics are written with low content and high redundancy or with impenetrably high-density content. So, it is a rare sign of pedagogical mastery combined with the genuine confidence of the experienced practioner when an introductory book manages to achieve a balance that is just right.

As I become more familiar with R, I still carry around this book in my briefcase for the occasional reread during which I uncover a nugget I had missed. When I have told this to my colleagues in computer science or bioinformatics, they immediately reveal that they share my enthusiasm for Dalgaard's work.

Let's be clear: this is a book that walks you through introductory and highly useful statistics while introducing you to the most effective ways to use R to perform these biostatistical analyses. It is not a programming book, nor is that its intent.

As an introduction to R, this book is very good. It's much clearer than the R documentation that comes with the code, and satisfied most of my needs. The statistical text was not very helpful, however. Discussion is very brief, and several points that would seem important are dismissed as beyond the scope of the work. I wasn't able to get a handle on the statistical tests I wasn't familiar with to begin with. The ideal audience for the book is people who know the stats already, and would like to learn R.

This book provides a very readable introduction to basic statistical analysis using R (with occational references to S-Plus). The table of contents displays the topics and I thought they were generally well covered in enough detail to compute the statistics (but this is not a statistics text). Especially helpful are the additional analysis steps, such as graphing results, and the peripheral R issues. Small things I would change: expanded coverage of manipulating data (e.g., SPSS's RECODE, TEMPORARY, MERGE FILE,...), more explicit instructions on installing the example data (it's at the end of the installation Appendix), discussion of interactions in ANOVA and regression, discussion of ANCOVA, and finally I would have liked a quick overview of the available packages and the stats they provide. But these are small issues; it's a great book.

This book seems like an excellent reference if you read though it in order and follow along using the example dataset provided online. However, I find that the transition to using my own data is far from clear. The book does not prepare you very well for using your own data, and barely discusses any type of matrices. This book if for univariate analysis, and univariate data.
As a reference it is definitely not as well suited. Looking up a topic in the index and jumping to that page often drops you in the middle of an example and you have to go back to the beginning of the at least the chapter to understand what's going on.
Basically after using this book I have found that though I can parrot the examples on the page, I do not understand the reasoning behind the code. Therefore when I go on to try and use my own data, I do not have the understanding and vocabulary necessary to adjust the commands to the needs of my dataset. For example there is a section on graphics (chapter 3) but I was not able to use it to help me figure out how to label a simple chart, unless it was the same chart that was used in the example.

I purchased this book after wading through the R help files for what seemed like hours. I was looking for a simple and straightforward guide that I could refer to for the basic operations of R. I am currently teaching myself C++ and learning how to interact with the Unix environment, but have very little experience with statistical programming. Therefore, I was looking for an accessible reference to help me become more comfortable with the R environment. *Introductory Statistics with R* provides concise answers to the "new user" questions that inevitably arise when programming in a new environment. In addition to its role as a programming resource, Dalgaard provides very useful information about the statistical methods he describes. I find this feature very useful as well, as I can rarely recall all the details of various statistical procedures from memory. My only caution is that this is an *introductory* guide to R. You will not find instructions for most (if any) of the additional libraries available to R. That said, I highly recommend this book to anyone who is interested in learning how to use R for statistical analyses.

I found this book very readable and a great reference for getting started with R. I was quickly able to run various tests from chi-square to logistic regression using this guide. I would agree that this would not be good for someone familiar with R, which is why it's called "Introductory". It also serves as a handy reference, providing easy look-up for how to accomplish various common (and some not so common) statistical tests.

The book provides good examples of how to run various types of functions and models in R. It could, however, better address the issue of how to import data from large pre-existing datasets into a format R can read. If an external program, such as Sed or Awk, is needed to prepare common file structures (rectangular or hierarchical) beforehand, it would be nice to know this up front. The book lives up to its claim that it provides a subset of information that R beginners may find useful and that it is not a definitive guide. The reader should beware, however, that while it gives a fairly thorough description of built-in R functionality, it may not be sufficient.

If you are new to statistics or have a limited knowledge of basic programming skills this book is not for you. If you understand basic statistics and know something about programming then this is an excellent introduction to how to use R to perform basic statistical analysis. It is not an R manual, as was stated in the preface. Nor is it an introductory statistics book by itself. It describes the analysis technique in high level, walks through the analysis step by step, and shows you how to use R to do the analysis. The chapter on linear models, specifically where he designs the matrices and dummy variables was a bit confusing. That was the only issue I had with the book.

If you have learned the stat concepts behind these procedures already, this is book is great. I think this book is very helpful for people who have had a few stat courses where the professor used a competing stat software package.

I really like how the author repeats some of the important syntax explanations throughout the book (for example, why attaching a data frame makes the syntax more concise). This approach makes it better than trying to learn R from the isolated html help files.