My "Idealism", isn't mine, it is the conclusions of researchers, I am not one, but I have made a point of reading a lot regarding campus violence and sexual assault, due to the situation we have. Regarding your "Per capita" as it applies to sexual encounters, or to how much and how often a given student drinks, I don't know how that would be done, other than questionnaires. I may be wrong, but I think a few responders may exaggerate their sexual prowess-just a little. But the police reports are hard and fixed, they can be counted. While we know, almost all of the claims of rape or sexual misconduct, or domestic violence reported (charged) against athletes is almost always false, usually just money grubbing whores. Nonetheless, these university studies are available to all with access to a computer. Don't attack me, attack the lack of knowledge and professionalism some of the University professors/academics show with their research.
Your example regarding Wal-Mart, the statistical accuracy IS a result of other factors, and those could/should be addressed once they have been ID'd. Just because there are other factors, in NO way makes the statistical probability any the less. This is just common sense.
There is a much higher probability you will be a victim of a crime(s) in various parts of large towns. This is a fact. The reason(s) are not addressed by that statement, it is a fact. For reasons why the statement is true requires more data. But, it is still a fact.