Lott is at it again. In a Tech Central Station column he claims:
Over 90 percent of the time simply brandishing the weapon stops an attack.
I suppose we should be glad that rather than his 98% estimate based on a fictional survey, or his 95% estimate based on a survey that gives a different number, Lott is now advancing a number that actually comes from a survey that was really carried out. Unfortunately, his 90% number is based on a sample size of just seven gun users. This sample size is too small to produce any meaningful estimate, and it is dishonest for Lott to keep presenting this number over and over again as if it meant something.
I appreciate that some people might find that the controversy about Lott's cooking of his "More Guns, Less Crime" results too complicated to follow, but this example leaves no wriggle room for Lott apologists.
- Lott's 2002 survey had just seven defensive gun users. Anyone can check this by downloading the survey data from johnlott.org
- A sample size of seven is far too small. Anyone who doesn't know this already can check this by consulting a statistics text or someone who has studied basic statistics.
How can anyone with a scrap of integrity excuse this?
Lott also deliberately omits any mention of the research that found that safe storage laws were associated with a reduction in juvenile accidental gun deaths. I suppose that is better than what he did in his book and blog which was to misrepresent the methodology of that research.