An absolutely brilliant little post by my colleague Philip Jenkins on the perils of relying on statistical evidence.

Alan Jacobs @ayjay