as an example, but after doing some research on the web, it seems a bit skewed.
I like your analysis better, but I think it is flawed as well, from a statistical perspective.
In your far right column, you show %REP-DEM, which is the % of states that went for Bush verses Kerry. While this analysis would be a correct method if all 50 states and D.C. had the same populations (and hence the same E.V.), the populations of the states vary greatly from state to state. Saying that California and South Dakota should be weighted the same when California's population is 46 times larger than the population of South Dakota is not a good assumption. Using your method greatly overstates the intelligence of those who voted for Bush and likewise understates the intelligence of those who voted for Kerry (since Bush won lots of small states and Kerry mostly larger states).
When you weight each state for population (
As you can see, after weighting each state for population, there is a much stronger correlation between IQ of a state's population and voting democratic than your analysis suggests.
Since it appears that you want your site to show as accurate representation of the data as possible, I thought you'd be interested in this.
Thanks for your time.
Sincerely,
Doug Waage
Sloan responds: Doug, thank you so much for your timely, useful chart. It now is featured on my new