re: #440 b_sharp

My point is that you can come up with insignificant correlations with almost any two events within the same domain. Even without worrying about causation which as you say takes a lot more than simple correlation, significance has to be measured at a reasonable level before you even worry about linking cause.

Fair point. It can be taken further, I think. Mass shootings are rare events, so it will be very difficult to show statistically significant correlations of such shootings with anything else because the number of data points will be small.