Gun ownership rate is a sticky thing, because there’s no direct measure of it. I used that data set because it was in a dozen big media articles at the time I wrote it.

Most academic studies use the ratio of gun suicides to total suicides. But when you use that data set, you get a similarly uncorrelated mess. You can try it yourself. This guy did:

Lying with Gun Data. Again.

I even tried to help him do so.