Try CNN’s exit polls here if you haven’t already. The vote by education figures are:
VOTE BY EDUCATION
BUSH KERRY NADER
TOTAL 2004 2000 2004 2004
No High School (4%) 49% +10 50% 0%
H.S. Graduate (22%) 52% +3 47% 0%
Some College (32%) 54% +3 46% 0%
College Graduate (26%) 52% +1 46% 1%
Postgrad Study (16%) 44% +0 55% 1%
BUSH KERRY NADER
TOTAL 2004 2000 2004 2004
No College Degree (58%) 53% n/a 47% 0%
College Graduate (42%) 49% n/a 49% 1%
The margin of error isn’t specified, but I think it’s ±3%. This would mean that most of the above figures are within the margin of error, and so there is no real correlation between whether or not someone is a college graduate and who they voted for. The only figure outside a ±3% margin of error is for those with postgraduate study, who tended to vote for Kerry.
Don’t forget that education is not independent of several important variables in an analysis like this, such as race and socioeconomic status. Just going by education paints an oversimplified picture.