I am not too sure this has a lot of weight:

When I was in college there are far more might owls than early morning risers. Out of a 100 people 80 are probably night owls (some of them might just end up partying the whole time). So there is an inherent bias there, because there are more nightowls than 'early morning risers'. Which leads to the conclusion in a sample of 100 if 10 of them are intelligent: then obviously 8 are night owls. I would personally never judge a person's intelligence on this factor (even though I am a night owl myelf).