Reply #237: There is no such thing as a "one-way correlation".

[DanTex]

Also, yes, the study did measure how likely people who did not get shot were to be carrying a gun. That's the "control" part of the "case-control" study. The results of the study are based on comparing people who did get shot with people who did not get shot, and adjusting for other potential confounding factors. And yes, you can "reverse the correlation" as you put it, provided that you are dealing with a rare event like getting shot, because, as I pointed out in my last post, in this situation the odds ratio will give a good approximation of relative risk. It's just arithmetic.

In any case, believe what you want. If you are actually interested in learning about the methodology in more detail, as opposed to insisting that you have uncovered a fatal flaw in this basic, widely used, and well-understood study design, then maybe some day you will get around to googling "case-control study" and reading about it. If not, then just go on patting yourself on the back.

Sorry if I come off as condescending, but what I am saying here is correct. If your college major was actually math, then you will be able to quickly read up and understand how it works, because in the end it is not all that complicated.