billvon said:
No. I did not use the Brady scores for any averages, just for the ranking of states with strict vs lax gun laws.
Any ranking that fails to agree with the plain facts posted - the actual laws, described and right there to observe - is problematic. Any ranking that did agree with those facts would also agree with the simple observation - the bottom five States in violent death have significantly less strict gun laws without exception, the top five in violent death include at least one State - Maryland - with significantly stricter gun laws. Two if one includes DC, which of course one should, as it is full of US citizens and has both its own gun laws and violent deaths.
billvon said:
I agree, though; if you look at just the top five and add DC you see the opposite correlation, primarily because you are skipping Hawaii (which is very low in violence and has a lot of gun laws) and are adding DC (which is not a state and passes laws in a very different manner from states.)
If you drop DC from the table (I didn't "add" it, it was right there) for some reason (your reason makes no sense - the manner in which the laws were passed is irrelevant here), and expand the example size to include Hawaii (Hawaii was not "skipped" - I chose five on each end because it came out to the extreme 10% on each end, which would be the source of greatest statistical significance), you still have no visible positive correlation between gun law severity and violent death rates. That contradicts your assertion, and supports mine.
I'm ignoring the still visible negative correlation, because I don't think gun laws actually correlate with non-gun violent death at all, in either direction - I think the negative correlation visible is an artifact of cultural features of the US population (the authoritarian hangover from the Confederacy and racism, the Asian influence in Hawaii, the farming and hunting and trapping practices of the near-Canadian States, and so forth).
billvon said:
I did! And the more states you use in your averages the more clear the correlation is.
That would be you, not me, using "averages"; you, not me, relying on dubious numerical scores for gun law severity that do not appear to match descriptions of the actual laws.
I'm using rank correlation, conservatively and prudently restricted to the extremes of what is in the middle a pretty subjective ranking system not suitable for fine discrimination. (The precision of the Brady "score" as you post it is reason enough for automatic rejection of its validity, even if it didn't fail to match observation of the data.)
The only correlation between gun law severity and overall violent death that is visible from ranking the actual gun laws themselves is negative. Expanding the rank sample to include the less statistically significant muddles the issue (it becomes much more difficult to rank the States in the middle), but we can easily see that places like California and Illinois and the rest of the Western States like Utah are going to reinforce, not invert, the strong initial pattern.