Statistics

My university training is as an econometrician, so I've had more exposure than most to the intricacies of statistical analysis. For many people the very word "statistics" triggers an instinctual flight response, but since you've clicked on the link to come this far, I'm assuming you're not amongst that number.

I remember a quote from a lecturer describing statistical modelling as "torturing the data until it confesses". That seems as good a definition as any.

Sites to Visit

• Statistics.com has probably the broadest coverage of all things statistical on the web. It links to software, jobs, data sources, courses and books, though with a distinctly US-centric viewpoint. The link to online books is especially valuable.
• If you're looking for free-to-own or free-to-try software, see the Freebies page from the comprehensive Interactive Statistical Calculation Pages website. This site will link you to hundreds of other sites at which you can perform various statistical tests online. A great example is the VasserStats website.
• Check out the R package, which is modelled on the commercial S-plus statistical package. The R package is free, and the base application is regularly updated and improved. As well, it's very extensible and individuals are constantly writing packages to perform specific functions. R is command line driven and takes a while to learn, but it's very powerful.
• There's an online statistical textbook at StatSoft.
• If your interest is in econometrics, specifically in econometric software, check out The Econometrics Journal Online.
• Online simulations to demonstrate various statistical concepts are available at The Rice Virtual Lab of Statistics.