Last week Alceste mentioned, in a comment about using personas for testing, that the sample set we have on www.microsoft.com/evidence may be self selected and skewed toward the educational market. That's a good point. If we had thousands of case studies,
