Paula, you wrote [in part]: > Now SAS certification. First, I don't know how many have already got it. > Then, If I interview somebody, how should I inteprete it? When typing this > msg here, I have three resumes in front of me. One says she has certificate. > The other two don't.

That's a hard problem this early in the certification issue. Lots of SAS coders do *not* have certifications yet. I believe that all the SAS Quality Partners now have to, and I don't think the re-testing is much more onerous or expensive than for the MSCE.

I think you can assume that the person with certification is ready to code in SAS, although the same may be true of the other two. But this won't tell you anything about their stat skills. See below:

> Somebody told me that if one is a very good applied statistician, the chance > is the person will be pretty good at SAS. But not vice versa.

I'd go further and say that neither may be true. Most SAS coders are not statisticians [and many of them try to stay as far away from stats as humanly possible]. But nowadays most applied statisticians are likely to be working in S-Plus or GLIM or R or any of a dozen other stat-oriented tools, rather than SAS. I do multiple languages myself, but that's because I can't get everything I want in a single package.

If I had to guess, I would say that

Prob(good at SAS given good at stat) is greater than Prob(good at stat given good at SAS )

But I'm a statistician, so I may be biased on this one. :-)

David -- David Cassell, OAO cassell@mail.cor.epa.gov Senior computing specialist mathematical statistician