Statistics is a very practical discipline, concerned about real problems in the real world.

A very well-designed and well-constructed text

An excellent online textbook

This online introductory statistics text is a companion site of VassarStats: Web Site for Statistical Computation.

an extensive web-based handbook of engineering-oriented statistical techniques. Closely integrated with Dataplot and other statistical software packages.

by Julian L. Simon. An introduction to probability and inference from a resampling (simulation) perspective. Covers probability, hypothesis testing, confidence intervals and sample size calculations and is an excellent introduction to simulation, bootstrap methods (sampling with replacement from the observed data to assess variability of a sample statistic), and permutation tests (comparison of observed differences between samples to differences obtained by randomly shuffling the samples together and re-dividing them).

an excellent introduction to statistical concepts, presented in a clear and non-intimidating way. Topics include: mean, median, percent changes, per capita and rates, standard deviation and normal distribution, margin of error, confidence intervals and sample sizes. Site includes a set of frequently asked questions, and a Statistics Bulletin board -- Post your question about statistics, or help someone else by posting an answer to his or her problem. Also links to its sister site: Finding Data on the Internet

a condensed handbook, or an extended glossary, written in encyclopedic format, covering subjects in statistics, computing, analysis, and related fields; meant to be both introduction and reference for data analysts, scientists and engineers (from CERN)

by Grinstead and Snell (a complete book, in Adobe Acrobat and Postscript format, that can be downloaded all at once, or chapter by chapter)

by Gary McClelland -- an online textbook containing many animated (Java) demonstrations. A tour, and part of the complete book, is available free.

giving the formula, expressions for the mean, SD, percentiles, higher moments, random variable generation procedures (if available), and insight into where that type of distribution function arises. In Adobe PDF format.

Philip B. Stark
Department of Statistics
University of California, Berkeley
Statistics means never having to say you're certain