<p>I looked up the report cited in post #725 that repeated this: </p>
<p>“Based on survey data from a 2000 study published in the Journal of Quantitative Criminology, U.S. civilians use guns to defend themselves and others from crime at least 989,883 times per year.”</p>
<p>I first looked up the source for that. It is a pro-gun website called “justfacts.com”. I then searched and looked the original paper, downloaded and read it. The paper says nothing at all like what justfacts says. Not even in the same universe.</p>
<p>The authors go through the two main groups of studies, some that show relatively low rates of firearm use in defense, others that show high rates. They identify possible sources of error and then describe a method which they propose to see which kind of error may be most important. They then run a survey. I can’t paste in the survey result tables or their discussion. </p>
<p>One result: they find the number of gun offenses, meaning criminal use, exceeds defensive use by a large factor. They noted another study’s estimate of a factor of 10, meaning 10x more criminal uses of guns. </p>
<p>Another result: they found some of the results were actually criminal. In other words, surveys that found high use involved actions that were actually the use of an illegally carried gun. So the finding is that people who have illegal weapons tend to use them - or brandish them - when they’re in the kind of situation that people with illegal weapons might find themselves. This suggests that use by law-abiding citizens in self-defense is lower.</p>
<p>And more: they found many of the uses were ambiguous at best. People seem to have guessed about intentions in interactions. We might want to believe the use of a gun comes from a guy coming up at you with a weapon of some sort but that’s apparently not the case. So people pulled guns - or maybe even said they had a gun - when no crime was intended. </p>
<p>And then: in terms of numbers, about 95% reported no incidents no gun incidents and 99% reported no crime incidents. This brought up the problem of valuing responses when you have relatively few positives. The authors note this may have skewed results in analyzing surveys.</p>
<p>As to the numbers cited in the quote above … they don’t exist in the paper. My guess is this website has taken some of the estimates referred to from other papers - without tracking those down and reading them - and then boosted them. Why conclude that? Because the next footnote in the document where this quote is from discusses how they adjusted the numbers in that case. I don’t have any interest in taking apart their statistics but they boost the numbers by extending them without much statistical thought to larger populations. </p>
<p>So the quote is what i would expect: a fabricated pile of dung. The moral: don’t rely on fabricated dung.</p>
![http://en.wikipedia.org/wiki/File:Percent_of_Media_Citations_by_Ideology.svg[/url]](http://en.wikipedia.org/wiki/File:Percent_of_Media_Citations_by_Ideology.svg%5B/url%5D){kind=link}
