It is interesting.
http://poseidon01.ssrn.com/delivery....087104&EXT=pdf
A quick scan suggests to me that their data do seem to support the idea that people of higher intelligence tend to hold views more consistent with their religious beliefs. Not sure that proves what they say it proves, there's certainly other possible explanations.
For example, instead of taking the data as suggesting something to the effect of 'smart people are better at talking themselves into believing things' (which sounds a bit like saying the smart people are outsmarting themselves), the data could also be interpreted in an entirely different way.
What the data might suggest is that dumb people are less able to see, or are more willing to accept, the inconsistencies of different beliefs they hold. The smarter people might have a world view in which it makes no sense to believe in both God and evolution, so they tend to choose one and reject the other. Conversely, the dumb ones either aren't as good at seeing the inconsistency in mismatched views, or for some reason aren't as bothered by it, as the smart ones.
After all, isn't part of being smart being able to identify when things do or don't add up? Isn't a dumb person more likely to believe two things that are inconsistent with each other?