I always had the impression they were made up by some journalist 10 minutes before the paper's deadline, ie so ad hoc to be useless for any sort of analysis.
I would be very interested in the level of data support you have for this correlation, Mark, sounds interesting.
Keep the good work up, Testa.