[WS:] That is a research question that can be addressed by introducing gun ownership (or perhaps prior gun training) as an intervening variable. So you would have two experimental groups: gun owners with access to gun and non-owners with access to gun, and one control group - folks with no access to gun (which could be both owners and non-owners). Issues like the one that you raise can be addressed by a careful experimental design. My point was not to produce such a design but rather to strengthen the point that using statistical correlations is not the best way to approach this subject.
Honestly, I am not sure what results would such experiment yield, but my suspicion is that most people would probably behave in a way they do in everyday life - they do not shoot people simply because they have access to a gun. But then, Milgram did not expect that the majority of his subjects would electrocute another person either. However, I also believe that people with emotional problems - and there is no shortage of those in our society - could be provoked and in this case the absence or presence of a gun would mean the difference between a black eye and death. My wife, who is the head of the ED department in a high school tells me stories about her students getting into vicious fights on the drop of a hat - black eyes, broken teeth, concussions, bitten off nipples, torn earlobes - one can only imagine what it would be if these people had access to weapons of any kind.
This is the case for a thorough background checks - which the NRA also opposes - but that is another story.
-- Wojtek
"An anarchist is a neoliberal without money."