« on: December 13, 2006, 09:20:53 PM »
Here's a hypothetical: Say the producers of the cream only surveyed the four dermatologists they thought were most likely to give them a favorable review. Maybe the cream gives you terrible pockmarks, but these four dermatologists are on the record with their highly unorthodox view that pockmarks are healthy in the long run b/c they create lift, like a golf ball's dimples, and preclude risky sexual activity. In that case, the producers might be able to get the 75% figure by interviewing only those four, when only about 1% or less of the total dermatologist community actually approves. In this case, the fact that it "isn't a representative sample" is an evident, serious flaw.