In a message dated 12/16/04 4:12:28 PM, [EMAIL PROTECTED] writes: >The correlation between per-capita state income and Kerry vote >percentage is +.70. That makes Bill by far the most accurate of our >guessers. If you do a bivariate regression, every +$1000 of per cap >income is associated with +1.48 percentage points of Kerry share.
I didn't think of placing a specific number on it until after I read Bill's estimate of .40. Then I couldn't decide between .25 and .50 and had to run to the vet. If we take my expected value as .25*.5 + .5*.5 we get .375--so Bill remains closer. :-) David