Thursday, January 13, 2011

Gun Frenzy: when statistics lie

I support gun control. I am firmly and proudly on the left concerning this issue. But I do so honestly. In the wake of the Arizona shooting, there has been widespread reporting of the rise in gun sales 60% since the tragedy. It's speculated that people are afraid of new gun control laws or perhaps some bizarre reaction to the publicity of weapons in the news.

But here's the problem:

In Texas , the number of background checks to purchase a gun rose from 621 to 724

nationwide it went from 7522 to 7906.

When it comes to the actual weapons sold, say in Arizona:

In the days since the tragedy, gun sales have increased dramatically. According to figures obtained from the FBI by Bloomberg, some 263 handguns were sold in Arizona the day after the killings, a rise of 60% on a year ago.

So you can do the math, we're starting out at a small number, and ending up at 263. A change in small numbers can produce big percentage shifts, that if reported without context are misleading. Going from 1 to 2 is a 100% increase. From 2 to 3 is 50%. From 3 to 4 is 33%. Yet each is an identical increase of 1.

If there's a 60% increase for a sustained period, say a year then it would be a legitimate story. Right now, 100 or so extra people in Arizona bought a gun. If 100 per day for 300+ days do it, that would be 30,000 gun difference; and a valid story. As it is right now, 100 extra people out of millions in a gun crazy state, after all the publicity is not a news-worthy story. If it is, use the hard numbers:

"100 more people bought guns in Arizona on this particular date than last year".
not the same impact is it?

No comments: