The Talkstats Bookclub


Global Moderator
Hi Guys,

What would you guys think of a talkstats bookclub?

The idea behind it would be this, we select a book, then every two weeks we review a chapter of that book here in the forum. We say what we like, what we dont, what we learned, we discuss things we are unsure about, answer each others questions and generally learn new things. Everyone who joins in on a book, is then expected to join in on the discussion (once every two weeks) and read the chapter.

My first two selections would be:
A) The R Book by Crawley (advanced chapters, does everyone have the PDF?)
B) The R inferno

Any other suggestions are welcome.

I will take the lead in organizing the first session & chapter, then a new member can take over.

I would like to know what you guys think.



Ambassador to the humans
I'd be up for it. I don't have the R Book though so I probably wouldn't participate there - unless I could find it cheap (very cheap).
I just bought it (and am trying to learn a lot more about R) so I would be interested. The book is BIG though so I think selecting a subset of chapters to go through would be best.


Ninja say what!?!
I'd be down if I can get a copy for free or very cheap.

To make it even more motivating and formal, why not set a time when we can meet online and chat? Then, I'd have even more reason not to put it off.


Probably A Mammal
A) The R Book by Crawley (advanced chapters, does everyone have the PDF?)
Whoa ... there's a pdf?! WTF am I doing carrying this book around?? lol

I totally like this idea. What chapters did you have in mind for The R Book? Depending on the aim we're taking, I think The R Inferno would be a better start, if our goal is to truly understand the inner workings of the R language; that book is amazing and insightful! Crawley's book would probably be better if our interest lies in learning more about using R for analysis, though.

Oh, and for another R book suggestion, for those interested in more numerical methods: Introduction to Scientific Programming and Simulation with R.
Last edited:


Probably A Mammal
You can put me down for the R book since I currently have it checked out for the next few weeks. I hate my UCD resources (used to be an econ grad there, and they never discontinued my resources! lol), since they are hard to manage. I don't know if I can get access to the pdf version. If anyone wants to not send it to me, PM me ;)

I've never used skype, but I'm sure I can pick it up. The problem with a specific meeting time is that we all have to be available for that meeting time when some of us may be on the other side of the world! We should start discussing if there is an appropriate time for a meeting. That is one of the benefits of the forum: we can communicate over time. Btw, skype is still free, right??


Global Moderator
I also believe that the different time zones would /could pose a problem especially for skype. Using the forum would make that less of a problem.

If we use the forum the rules will be, you have up to a specified date (14 days after the start date) to read and discuss a chapter. After that date, we will move on to the next chapter and the thread will be closed - so you have to be on time to contribute, and discuss or get any answers to questions you have.

TS Book club thread should start with the tag [TSBC] specifying TalkStats Book Club and then the book title and chapter.. e.g. [TSBC] The R-Book, chapter X, regressions.

Oke time to select the chapters:

Getting Started
Essentials of the R Language
Data Input
Classical Tests
Statistical Modelling
Analysis of Variance
Analysis of Covariance
Generalized Linear Models
Count Data
Count Data in Tables
Proportion Data
Binary Response Variables
Generalized Additive Models
Mixed-Effects Models
Non-linear Regression
Tree Models
Time Series Analysis
Multivariate Statistics
Spatial Statistics
Survival Analysis
Simulation Models
Changing the Look of Graphics
As this is not going to be a beginners course I suggest we skip everything up to Statistical Modelling, and make a choice from there. I would be interested in GAMs & ME-models. Although i think we should start with a simple chapter as regression or GLMs so the threshold is not too high. I'm likely to learn from any chapter we try.
I thought numbering them might be easier

1	Getting Started
2	Essentials of the R Language
3	Data Input
4	Dataframes
5	Graphics
6	Tables
7	Mathematics
8	Classical Tests
9	Statistical Modelling
10	Regression
11	Analysis of Variance
12	Analysis of Covariance
13	Generalized Linear Models
14	Count Data
15	Count Data in Tables
16	Proportion Data
17	Binary Response Variables
18	Generalized Additive Models
19	Mixed-Effects Models
20	Non-linear Regression
21	Tree Models
22	Time Series Analysis
23	Multivariate Statistics
24	Spatial Statistics
25	Survival Analysis
26	Simulation Models
27	Changing the Look of Graphics


Ambassador to the humans
My vote goes for 13 - GLMs (I just love em' so much... but I haven't explored too much of R's capabilities because mostly we programmed all the stuff ourselves when we were learning about them)


Probably A Mammal
I'm finishing up my brief stint with logistic (and GLMs) regressions sometime in the next week or so. From there I'm focusing on ANOVA generally, but I am interested to know more about analysis of covariance. Therefore, my vote is not to ignore 11 or 12. When we get to them is a matter of taste. I have a lot of interest in 23, though. I just don't know if I have the requisite knowledge to fully appreciate it, yet. I also have a recent background in GIS, and I find 24 could prove useful to know. I also think 20 could be of great interest along with 23. As for 27, I think that should just be ignored for another book (e.g., the lattice or ggplot2 book, both of which I found very useful for learning visualization techniques). Also, 22 could be deferred to another book.


Super Moderator
Im in. Either book would be good for me. Thanks for the heads up on the RInferno Marco, I appreciate that. It looks pretty full on!

If we go with th R book, I'd be keen to discuss the GAM chapter also, but the mixed-effects models could be interesting too.