Your point is well taken: we can't just assume this stuff, and it really does have to be measured. And I'm not remotely an expert on this stuff, so I can't prov

Your point is well taken: we can't just assume this stuff, and it really does have to be measured. And I'm not remotely an expert on this stuff, so I can't provide thorough evidence for this. However, I've heard plenty of people who have studied this stuff much more than I have present this connection as being pretty well established by direct evidence, and given that the theory is generally plausible I've seen no reason to think they're lying. That being said, here's a link: http://psp.sagepub.com/content/32/3/286.abstract http://psp.sagepub.com/content/32/3/286.abstract At least from the abstract, it sounds like it's addressing exactly the question you've asked (and that previous studies have found related results). (I took that link from a recent blog post by Jim Hines, which includes quite a few others worth reading, too: http://www.jimchines.com/2013/03/preventing-rape/ http://www.jimchines.com/2013/03/preventing-rape/ .)