Deltoid

Steve D. Fischer writes:

While you're at it, keep in mind that one of Pim's favorite scientists (i.e. one who also hates guns), Colin Loftin, has said publically that the NCVS (National Crime Victimization Survey) survey - the "Gold Standard" (guffaw) of surveys - undercounts spousal abuses by as much as a factor of 12, and rapes by a factor of 33.

Err, no. He said it might undercount them by this much. Criminologists do agree that NCVS significantly undercounts non-stranger crimes. It does not follow that it undercounts stranger crimes. Since the uses reported to Kleck were mostly against strangers this does not explain the enormous discrepancy between Kleck and the NCVS.

If you've had to discharge your weapon and you've shot someone then the police have probably become involved and you have the incident on your record. There is no reason to lie about it when talking with an interviewer from the government.

The Kleck estimate of the number of times defensive gun users shot someone is more than twice the the NCVS for all DGUs.

However, if no shots were fired, the odds are that you didn't report the incident and only you and the bad guy know about it.

Not according to Kleck's study. 64.2% said the police knew about the incident.

You're not likely to admit that to an interviewer who represents the Justice Department, regardless of how much they might insist that your responses are confidential.

According to Kleck's study in 64.2% of cases the police already knew about it, so there seems no reason for these to hide it from the Census Bureau.

Secondly, you're naturally going to feel ill at ease admitting to someone sitting in front of you that you may have broken the law, even if it was to defend your life. People have a natural tendency to want to please the interviewer even if that person is a stranger. That face to face contact makes all the difference in the world.

NCVS interviews are mostly conducted by phone. The natural tendency to give a response to please the interviewer is a threat to both the NCVS and Kleck's study. Kleck wanted to study the nature of self-defense with a gun. Each additional piece of data would undoubtedly please him.

You can argue about how many people the NCVS interviews versus the other surveys until you're blue in the face, but it won't alter the fact that if 97% of the people you interview about certain victimizations lie, the size of the study is irrelevant.

And you can argue (as Kleck does) about how many other surveys give similar results and it won't alter the fact that if 5% of the people you interview about defensive gun use lie, the number of studies is irrelevant.