List of recommended stats books

#1
Stats novice here but enthusiast. What would be a list of books you would expect any BS in statistics to have. I have my list culled from uiuc's course pages (from what I could find). Any recommendations for someone trying to attain broad knowledge (or at least reference) in the subject?


using R for introductory statistics verzani,j chapman hall
introduction to the practics of statistics mccabef, g and d moore wh freeman
fundamentals of biostatistics bernard rosner
probability and statistical inference robbert hogg and elliot tanis
introduction to mathematical statistics hogg mckean craig
statistical models a.c.davison
elements of statistical learning: data mining, inferen e, and prediction hastie tibshirani freidman
plane answers to complex questions
introduction to categorical data analysis alan agresti
recommended: the little SAS book lora delwiche susan slaghter
modern applied statistics with S w.n. venables b.d. ripley
time series analysis and its applications with R examples shumway and stoffer
 

Dragan

Super Moderator
#2
Stats novice here but enthusiast. What would be a list of books you would expect any BS in statistics to have. I have my list culled from uiuc's course pages (from what I could find). Any recommendations for someone trying to attain broad knowledge (or at least reference) in the subject?


using R for introductory statistics verzani,j chapman hall
introduction to the practics of statistics mccabef, g and d moore wh freeman
fundamentals of biostatistics bernard rosner
probability and statistical inference robbert hogg and elliot tanis
introduction to mathematical statistics hogg mckean craig
statistical models a.c.davison
elements of statistical learning: data mining, inferen e, and prediction hastie tibshirani freidman
plane answers to complex questions
introduction to categorical data analysis alan agresti
recommended: the little SAS book lora delwiche susan slaghter
modern applied statistics with S w.n. venables b.d. ripley
time series analysis and its applications with R examples shumway and stoffer

Statistical Inference
by George Casella and Roger Berger is a good book to add to your list.
 

Dason

Ambassador to the humans
#3

Statistical Inference
by George Casella and Roger Berger is a good book to add to your list.
Indeed. I think the problem I have with this thread (not really a problem... but the reason I haven't said anything) is that I don't really feel there are any books that I would expect everybody to have. There are some books I could recommend... but I wouldn't call them essential or expect others to have them.

With that said I'll second the Casella and Berger nomination. Although I think it's primarily used as a masters level theory book I think undergrads could get quite a bit out of it. It's well written and has plenty of examples while still leaving enough to the reader to parse through to learn the material well.
 
Last edited:
#4
The problem I have with statistics is that there seems to be so many different methods and subfields that it is hard to get the 1000-mile-up view of the subject (especially for a neophyte). I would love a book that just briefly describes the many different methods from time-series to multivariate analyses, to data-mining, to the stats used in paleoclimatology to the stats in econometrics. I'm sure that statistical methods are constantly being developed and are the subject of many-a-phd thesis, but I would devour one of these entire-field-skin-deep books. In lieu of such a book, I figure a range of books intended for someone with broad knowledge in the subject would suffice.
 

Dason

Ambassador to the humans
#5
The problem I have with statistics is that there seems to be so many different methods and subfields that it is hard to get the 1000-mile-up view of the subject (especially for a neophyte).
Is there some subject that doesn't have this problem? (Seriously if you think there is let me know)
 
#6
The one book I expect everyone to have is Gonick's The Cartoon Guide To Statistics.

I bought Agresti's book but have been pretty disappointed by it. It's mostly examples of results from black-box software algorithms with little mathematical justification and no discussion of algorithms. This is a pretty typical problem for most books of the "Statistics for Discipline X" variety. (Statistics books for econometricians seem to have escaped this curse, probably because econometricians are really full-fleged statisticians.)

My favorite statistics book is an oldie: Cramer's Mathematical Methods of Statistics manages to quickly get to the essence of concepts without leaving out the mathematical derivations.
 
#7
Is there some subject that doesn't have this problem? (Seriously if you think there is let me know)
Almost any of the Humanities. They even make overview courses, they call them 'survey courses'. But point taken, maybe I should start on a book series for the sciences ;). Actually there are great overview book SERIES for Economics, Physics, and Medicine (oh Medicine is FULL of them). Is there a similar statistical review series out there?
 

bryangoodrich

Probably A Mammal
#8
I'm also on board with Statistical Inference. I heard someone mention it over at MathHelpForums, so I checked it out from the library, and find it absolutely amazing. I'll add it to my library when I can afford a **** near $200 book! In my library is Kutner and other's Applied Linear Statistical Models. It was so comprehensive, and UCD uses it for two of their courses, I decided to work my way through it for the applied training my school lacked. Those are my two favorites at this time: one more theoretical and the other applied.

I always recommend checking out the syllabus used in classes at your favorite university. See what is actually being used out there, and then check out the Amazon comments to get further ideas or whatever else Amazon might recommend.

Oh, and one I also like (theoretical) is Ash's Basic Probability Theory. He does a great job explaining ideas, and since the book is an oldie, the mathematics is a bit more rigorous and in-depth than a lot of schools will teach at an early level. He also has a graduate level book on measure theory I want to look at some day, which also has great reviews reflecting on his ability to articulate ideas.
 
Last edited:

bryangoodrich

Probably A Mammal
#9
Oh, and I've used Hogg and Craig's book mentioned in the OP. It sucks! Unless you have the solutions manual or someone to help you along the way, that book is a useless resource for learning statistics. I'm sure it is great for reference to someone that already knows the material. I own the Data Mining one mentioned in the OP, too. It is way advanced for a novice, but the material is great. I saw some bad reviews on Amazon, but that all just sounded like whiny complaints ("they don't do this" or "they didn't show that").
 
#10
Seconding bryangoodrich's recommendation for Kutner et al., Applied Linear Statistical Models. It's a very readable book, and would be ideal for self-study (as well as being an excellent course textbook.) You could probably design an entire MS-level curriculum around just this book plus Casella and Berger.

The undergraduate equivalent, more or less, is Devore's Probability and Statistics for Engineering and the Sciences. This is the book I'd recommend for anyone from a math or computer science background who wants to get up to speed on statistics. There are also R packages on CRAN with all the data sets, which is nice.