don't bother<quoted text>
Who on Earth would rely on a study which used self reporting interviews to collect their data for something like this?!
The study I was thinking of (and the type I would be comfortable looking to in matters of statistics) is an empirical study.
Counting the numbers of assaults that resulted in shootings and then determining if the victim had a firearm. If you can access academic journals, I'm more than happy to find the title for you. It's got a pretty good breakdown of the methods.
ask Trayvon how your study worked out for him
PS: if I were thrown in a knitting contest...I would not fare well.