>Anybody done any statistics, that's willing to share it?
>
>I was curious about things like standard deviation, variance, tests for normal etc..

Don't know if this'll help much, but here's a few cents on this...

I've written specifically-tailored sampling/variance systems entirely in vfp - before that, I used C in older PC environment and PL/I on the mainframe to do similar. But vfp makes everything vastly easier with its table processing capabilities (like indexing, e.g.). Leaves C in the dust as far as coding ease, if not speed...you just need to get statistical formula/algorithm requirements set, and then just code them. The hard part really is making a good GUI for the users to do their work from inputting raw frame data, refining the frames, through selecting samples and refining the samples, etc, etc.

As Erik said, vfp has a few basic features like STD, which of course gives variance as well. But as far as tests for distribution, we use SAS. There's no sense in trying to do that sort of thing from scratch, when you have such complete tools like SAS or SPSS available. You start getting into matrix calculations that get nasty in a hurry, in many statistical calculations, and you need a lot of pre-built lookup tables, as well.

But, in theory, you could do the same in vfp if you wished to, maybe if you only needed one type of test, and had a budget constraint, for example.