Best books on statistics

www.iSteve.com/05FebA.htm#best.book.on.statistics

Question: Best book on statistics for beginner? A reader writes:

I'm looking for a good book on Statistics that can
take one from the basics to an advanced level while
building intuition. I was wondering if you had any
recommendations.

I received a couple of quick votes for Larry Gonick's Cartoon Guide to Statistics.

And a reader who is much smarter than me sent this list:

For self learning you need to have good books w/ worked examples and answers to even problems (at least) in the back of the book so that you can check yourself. My recommendations are made with this in mind.

I've read all of the following books pretty much cover to cover, so I'm not recommending blind (w/ the exception of Jun Shao which I keep around for reference purposes). With that in mind:

1) If he doesn't know probability, start with Pitman.

He should also know linear algebra to the level of Strang.

2) If he already knows probability, I like Degroot and Schervish.

Supplement with Schaum's if more practice is needed w/ specific problems.

This will get him to intermediate level. At this point he needs to figure out where he wants to go.

A) Computational statistics and data mining:

Duda & Stork

and *DEFINITELY* Venables and Ripley (which he can go through with R).

Depending on whether he's feeling his oats, he can also go for Hastie and Tibshirani, but it is somewhat opaque at times (though worth going through).

I also strongly recommend going through the various tutorials on the R project site. A lot of that stuff has not been put into book form but is very useful.

B) Stochastic processes, filtering, and time series - I like Papoulis

and then Oksendal:

C) Measure theory and theorem/proof stuff - go w/ Jun Shao. My background in this area is not as strong.

Well, we all have our weaknesses. Personally, I think I'll stick to the Cartoon Guide, myself.