American football and violent crime. More fun with Janet Hyde.

Here is something extra I had to comment on but just thematically didn't fit in my last post. Also on Joy Cardin's show her guest Janet Hyde and a caller decided to slam competitive sports. The caller suggested that competitive sports increased aggressive actions, as opposed to acting as a cathartic release of aggressive tensions in the athletes. Specifically they, the caller, and Professor Hyde believed that participation in competitve sports, football specifically, though they included basketbal-men's and women's-in the list of such sports, that could induce higher levels of aggression.
Janet rejects what she called the catharsis hypothesis based on what seemed like a single study that looked at rates of violent crime among football players. College football players I think. Anyway the conclusion, football, and other violent competitve sports encouraged, and increased aggressive crime. I am not sure what group to which these violent athletes were compared.
Lets not try to deny that football is a rough and violent sport. That would be silly. However gentle reader, you may see the same flaw in the study that I do. Its almost as bad as a self-selecting survey. People interested in football or some other violent competitive sport are by their very nature going to tend to be the more competive, more aggressive than other students in general. So essentially the researcher is comparing different populations of people. Competitive, risk taking people, to the rest of the school at large. It isn't a random sample. That is to say, the researchers aren't randomly selecting from a large group and subjecting them to a program and then looking at the results.
The only way to prove the hypothesis, playing football or (any physically competitive sport) increases the violent tendencies, of the males and females who play them is as follows. At several schools you'd have to take a random sampling of students, have them play for four years or at least two, and then compare rates of incidence of violent crime with the non-sport playing groups. That certainly wasn't done. Also it would be nice to compare to these groups (sport playing group, to non-sport playing group) to non-experimental sports groups (by that I mean teams not from your experimentally designed programs. And you if you were looking at any factor other than simple rate of violent crime, say incinidence of aggression say, you'd want the research to be run more or less double blind.
Now I've helped some promising grad student under Professor Hyde design their Ph.D research.