Tuesday, January 16, 2007

Harvard's "Study"

You may have heard about Harvard's anti-gun "study." Go over to Alphecca's response. It's good. Thanks to Blogonomicon for the link.

While I disagree in principle with taking away guns whether they cause or prevent crime, it is always good to show that they don't cause crime. After all, that's one of the big battlecries in the push to ban guns: "But less guns will mean less crime!"

And, as some have noted, the study is flawed because of several things. A lot of gun owners are a bit shy about telling some stranger on the phone that they've got guns. The study adjusts for many factors (which begs the question of the relevance of these factors), but asks a random sampling from each state in question, which can easily lead to crime rates in urban areas being linked to gun ownership in rural areas. The study also polled only 200,000 people across all 50 states. The current estimated US population is 300,983,000 and a few hundred in change. The sample size, then is less than 0.07%. should have provided us with some plus-or-minus adjustments to the results. I can't afford to buy a $30 report, but no chart I've seen shows any statistical margins of error. This could be something I would find in the full report, so don't take my word on it. Finally, it was backed by a noted anti-gun organization, which calls into question the validity of the study (Geek with a .45 and The Smallest Minority have more on the Joyce Foundation). After all, the anti-gunners would never accept a study backed by the NRA.

Read Alphecca's response to the study. He's more thorough than I am.

No comments: