Reply #54: Holy shit. I think we almost agree.

[jgraz]

The real problem is that Kleck's methodology cannot be verified by independent means. And most attempts to correlate his results with other data do not hold up. For example, his numbers suggest a frequency of justifiable homicides that is around 10x the documented rate (quoting from memory).

With rape statistics, there are many ways to gather them. You can use official crime reports, victim surveys, perpetrator surveys and healthcare provider surveys. These data can be used to reinforce (or refute) each other.

Even with all the data, most rape studies admit the deficiencies with self-reporting -- especially the strong dependence on the ways the questions are worded. Kleck admits no such deficiencies, even though his results are not in sync with any of the other surveys he cites.

Kleck also is hardly an unbiased researcher. Read his study ( http://www.guncite.com/gcdgklec.html ) and you can easily suss out what his point of view is. It's hard to imagine a rape researcher having such a strong stake in the data coming out one way or another. And that stake can have a huge effect on the results of the survey -- you only need to look at Republican vs Democratic issue polling to see that.


Of course the biggest problem with Kleck's survey is it tells us nothing of value. He doesn't even attempt to compare DGUs with other means of deterring crime. DGUs are naturally going to be prevalent in the US because guns are prevalent in the US. But we have no way of knowing how many of the alleged DGU crimes could have been deterred through other means.