I've been looking at the data and graphics at the excellent "Graphical Representations" blog. While there are many things there worth spending some time staring at, I was particularly interested in his post graphics matter, international edition which uses data from the international small arms survey to correlate gun ownership with murder rate in nations.
Here's his resulting chart with trend line. (You'll want to click on it to see it larger, really you will, or go to the blog post above.)
Now, I have a problem with that chart, because it compares apples to kiwi to marshmallows. It includes developed nations with excellent police systems and legislatures and nations that are practically failed states, nations with active drug cartels and revolutionaries killing people.
But, he always links to his data, so I thought: "Lets restrict it to developed nations without international or civil wars going on in them, who aren't hosting pirates or drug cartels and have actual economies, albeit some are better than others...
And I excluded the US. I'm interested in the idea if in the rest of the developed world there is a correlation between civilian gun ownership and murder rates.
I freely admit that the list of nations I chose out of his is prejudiced and entirely arbitrary. Tough. My blog, my graph, go make your own.
R2 = -0.0002 I.E. the line is flat, there's no correlation.
Hmmmm. And - as always - Show your work