Nothing helps an argument like poorly assembled and decidedly misinterpreted statistics (when your opponent starts to point out your flawed reasoning, call him a Nazi). Still, Vali Chandrasekaran of Business Week thinks that we should all be careful drawing conclusions from merely correlated data points. There are several more humorous examples at the link.
Link -via @MarilynTerrell | Previously: The Science News Cycle
A great deal of bad policy - including economic policy - is created because of a faith in correlation=causation.
We are warned about this, but it's such a common mistake by most.
and furthermore
Statistics Science
It's amazing how many people confuse statistics and science, especially people who should know better.