Today, I thought I'd do a quick diary and call for participation from the other quant people around here.
I think there are a fair number of kossacks who, for one reason or another, need to learn some statistics. Some are in school. Some need to do research. Some wamt to read statistics.
I try to cover some stuff in these diaries, but diaries are evanescent, books are permanent.
OK I am not at my office, but will start this list now, others can chime in, I'll add more tomorrow, and as I think of it. Then maybe tomorrow or Tuesday I can make a better list. When you add a book, please add comments.
My focus is on social science type statistics, which ought to match with a bunch of users here, but feel free to chime in. I'm going to organize the list into a few categories, but feel free to suggest others
Update [2006-12-4 10:1:51 by plf515]:
Basic and Introductory and General and Miscellaneous
- Statistics as principled argument by Robert Abelson. A great book. Very easy reading. Very little math. This book won't substitute for a text, but is a great book for anyone who knows a little statistics. Rather than teach methods and formulas and such, it does what the title says: Show how statistics is a way of arguing. If you read statistics, or do research, you should have this.
- The elements of statistical learning by Hastie et al. This is NOT an introductory book, but I did not know where to put it. A very general book, and a beautifully produced one.
- Introduction to the bootstrap by Efron and Tibshirani. If you want an intro to the bootstrap, this is it. Bootstrapping books tend to either be a) So simple that they teach nothing or b) Meant for mathematicians. This one strikes a happy middle ground. If only they would update it!
- Classification and regression trees by Breiman et al. I think tree methods are very neat, and under-utilized. This is a good intro.
- Statistical reasoning for the behavioral sciences by Richard Shavelson. Recommended by teacherken
- How to lie with statistics by Darrell Huff recommended by dmsilev and coss and mgk.
- A mathematician reads the newspaper by J. A. Poulos recommended by mathguyntulsa and mgk
- Elementary statistics by M. F. Triola recommended by mathguyntulsa.
- Introduction to the practice of statistics by D. S. Moore and G. F. McCabe. recommended by mgk.
- Cartoon guide to statistics by L. Gonnick and W. Smith. recommended by virago.
- Intuitive biostatistics by H. Motulsky, recommended by mindgeek
- Probability theory: The logic of science by E.T. Jaynes et al., recommended by Alex in Osaka
Exploratory analysis
- Exploratory data analysis by John Tukey. An odd book, but very good. Pre-computer era.
Regression methods
- Regression modeling strategies, by Frank Harrell. This book covers a wide variety of regression methods, and attempts (fairly successfully) to give a general approach to analyzing this sort of data. Not an introductory book, but good if you've had a couple courses in regression and want to see how it should be done.
- Regression models for categorical and limited dependent variablesby J. Scott Long. For the times when your DV is not continuous.
- Applied logistic regression by Hosmer and Lemeshow. One of the clasics in the field.
- Models for discrete longitudinal data by Molenberghs and Verbeke. A good book on an advanced topic. Not for neophytes, but excellent. Especially recommended if you use SAS.
- Linear mixed models for longitudinal data by Verbeke and Mollenberghs. See 4, immediately above.
- Applied longitudinal analysis by Fitzmaurice, Laird, and Ware. In my opinion, this book is as introductory as you can get in this topic without losing all meaning.
- Statistical reasoning in behavioral research by E. J. Pedhazur
Graphics
- Visualizing data by William S. Cleveland Wonderful stuff. Cleveland knows more about statistical graphics than anyone else. He also wrote
- Elements of graphing data which also belongs on the shelf of anyone doing graphical work, or looking intently at graphs
- All of Tufte's work is good, although it is less practical than that of Cleveland. Tufte's books are beautiful.
Categorical data
- Categorical data analysis, by Alan Agresti This is the book for those who want absolute authority and fairly detailed mathematical background. Although it's not an easy book, it's hard because the material is hard; given the level, Agresti is very clear. If you want a more introductory book, see 2.
- An introduction to categorical data analysis by Alan Agresti. I haven't read this one, but have heard good things. There seems to be a new edition out soon. If you want something even more basic, there is
- Applied categorical data analysis, by Chap Le. This is not the final word; the author leaves out some key details. But it's a starting place.