(March 31, 2014 at 1:49 pm)Jacob(smooth) Wrote: Put it another way.Actually, no, that's not the case. Note that they don't say homosexual, which refers to an orientation, but rather MSM, which specifically refers to behavior. A man with homosexual orientation who has not had sex with a man in a long enough time would not be rejected.
Homosexual people are more likely to be HIV positive. We deduce that this is because they are more likely to engage in behavior which causes HIV.
Quote:Black people are more likely to be HIV positive. We deduce that this is because they are more likely to engage in behaviour which causes HIV.So why not ask about the behaviors?
Note that I said that having correlation with a known mechanism is more compelling than a correlation alone. I didn't say that correlations alone are worthless or that we can't act on them. Before the mechanisms of smoking and lung cancer were known, the correlations between smoking and cancer were worth action. In this case, if all we knew about HIV were that it was more prevalent in black people, then that could be actionable. But, since we do know about the transmission mechanisms, it makes more sense to use those.

