I'm looking for suggestions on a good statistics book.
For my book project, it's becoming clear that I need a deeper understanding of the statistical issues associated with climate research. Many of the data sets used are easily publicly available, so it makes sense to deepen my understanding by actual playing with the number myself. But to do that, I need a better understanding of the statistical tools.
I'm getting some help already from a volunteer "statistics coach", a friend who's a physician with a highly developed interest in the statistical problems of the practice of medicine. He's loaned me his copy of Stanton Glantz's Primer of Biostatistics, which seems promising. Its focus is on medicine, but much of it seems completely generalizable.
I'm looking at something that is pitched at undergraduates, ideally with a lot of practical examples/problems on which I can practice. As a bonus, a book with an on-line component would be great - downloadable data to play with.
Suggestions?
Posted by John Fleck at February 05, 2004 09:39 AMHow about this one. It's old, but still quite good: http://search.barnesandnoble.com/booksearch/isbnInquiry.asp?isbn=0393310728&itm=1
Posted by: Jonathan Blandford on February 5, 2004 10:29 AMDepends what kind of stats you want to do methinks.
For beginners I would suggest Howell (http://www.uvm.edu/~dhowell/StatPages/StatHomePage.html), for further exploration, there is another but I cannot remember the author (there are "student" and "lecturer" editions though) and "Nonparametric Statistics" by Segal and Castellan
For climate analysis, have you looked at CDAT, a Python-based open source set of climate analysis tools made by the folks at Lawrence Livermore labs? (ref: http://esg.llnl.gov/cdat/)
Sorry this stuff is biased towards psychology, but that's my particular field.