I'm going to point out that the point you
don't touch on is that Cook & Ludwig's observation was of their
own study, not of the Kleck & Gertz study, like you claimed. I surmise that you're hoping the readers will overlook that particular glaring piece of evidence of your inability to read what an article actually
says, rather than what you want it to say (or "gun-aversive dyslexia," as it has been called).
1) The critique was written by some guys who did their own study which, using a methodology similar to Kleck's, came up with even more ridiculous numbers -- numbers which the report you link to admits are highly suspect.
2). Some doctor (or maybe a guy from Maryland) had some snarky comment about the authors.
An ounce of evidence beats a ton of speculation. And speculation is all that Cook & Ludwig can produce in trying to explain away why their survey produced results that weren't to their liking. Woolley's question is not mere snark (your attempt to dismiss it as such notwithstanding), it's a valid question: if it's supposedly so obvious that the methodology used to produce these findings was ill-suited for the intended purpose,
why use that methodology at all? Cook & Ludwig must have read the Kleck & Gertz study, since they purposely set out to see if they could replicate its findings; if there were any glaring errors in Kleck & Gertz's methodology, Cook & Ludwig were equally in a position to identify those
prior to formulating their own protocol as they were
after having implemented it, so why didn't they avoid any such errors from the outset?
Moreover, Cook & Ludwig do not express doubt as the veracity of their respondents' replies on
any other aspect of their survey. In particular, they note that their data "indicate that just 35 percent (plus or minus 1.3 percent) of households own guns" and state that:
This estimate may be somewhat off the mark but not by much. Conventional wisdom <that something in the order of 50% of American households own one or more guns> appears out of date.
So Cook & Ludwig express an extremely high level of confidence that their estimate of the percentage of American gun-owning households is correct, even though it is inconsistent with the findings of numerous other surveys. That includes one study published two years earlier that found that 10.3% of respondents known to hold hunting licenses, and 12.7% of respondents known to have a handgun registered to them, denied owning firearms when interviewed (Rafferty, Ann P. et. al. "Validity of a household gun question in a telephone survey."
Public Health Reports. May-June 1995 v110 n3 p282(7)).
So if, even in the face of this contradictory evidence, Cook & Ludwig are
so confident about their respondents' veracity when it comes to firearms ownership, why does that confidence completely evaporate when it comes to the number of reported DGUs? Could it be that the finding that gun owners are a minority, rather than a plurality, in American society conformed to their personal agenda, while the finding that annually more DGUs occur than gun crimes did not? Naaah, perish the thought!
Also, would you care to explain on what basis you consider it "ridiculous" that in a country with (at the time of Cook & Ludwig's survey, i.e. 1994) a population of 265 million, divided into 97.1 million households of which at least 35% (~34m) but possibly as much as 50% (~48.5m) owned a firearm, a member of between 3% and 4.4% of those gun-owning households experienced a DGU?
3) You think you might recall some different arguments I used at some time in the past. You would prefer to argue against those statements rather than address the topic at hand.
Considering you list six other items in your post, suggesting that I am somehow failing to "address the topic at hand" is really lame. You've seen that it's not the only thing I wrote
because you responded to the others. How disingenuous can you get?
Anyway, I'm just anticipating you pulling out your argument that "Cook & Ludwig produced findings that replicated Kleck & Gertz's results because they used Kleck's flawed methodology" that you've used before, mostly to save typing later. I really should stop tipping my hand, I acknowledge.
4) Two other guys conducted a study which produced a result almost three times lower than Kleck's, and that somehow supports his findings.
Sample-based statistics isn't an exact science, because so much depends on how representative your sample is (and the only way you'd know for certain is by polling the entire population, which is practically impossible, which is why we have sample-based statistics in the first place). Kleck & Gertz's confidence interval was fairly wide; they reckoned that the actual number of DGUs could range from 1.5 million to 3.6 million, and that's based on a significantly larger sample (4,977) than that used by Cook & Ludwig (2,568) or Hemenway & Azrael (1,905). The larger your sample, the more likely it is that your sample is representative of the general population, and the narrower your confidence interval. Because G&L and H&A's samples were smaller than K&G's, the margin of error is much larger, and it is therefore entirely plausible that the actual number of DGUs falls (or at least fell during the 1990s, since violent crime rates in 1993 were more than double what they are now) into the 1.5-2 million range; at least half again as many, and possibly twice, the number of reported violent crimes committed using a firearm at the time.
Moreover, it should be noted that David Hemenway and Deborah Azrael are not simply "two other guys" (for starters, the name "Deborah" should give a hint); they are two of the most prolific crankers-out of public health studies that (invariably) conclude that Guns Are Bad (along with Arthur Kellermann, Garen Wintemute and Matthew Miller). When two of these luminaries of anti-gun journal publications produce a study that concludes that there are more DGUs annually than there are violent crimes involving firearms, that's kind of significant. Like Cook & Ludwig, they immediately tried to rationalize why their methodology was all wrong for the purpose of producing accurate results, but if we accept that, we have to call into question the findings of all their other studies. After all, if they--like Cook & Ludwig--are by their own admission too incompetent to come up with an adequate protocol in this instance, why should we assume that any less incompetent the rest of the time?
The interesting thing is that these supposedly invalid studies (only found to be invalid after the fact) are quite consistent with the findings of over a dozen other studies conducted prior to the Kleck & Gertz study, the main difference being that those studies were not set up to specifically ask about DGUs. Replicability of results is the best available indicator of any scientific study's validity, and the fact is that both Cook & Ludwigs' and Hemenway & Azrael's results are consistent with the existing body of knowledge. So we have a large amount of evidence that there are
at least several hundred thousand DGUs in the U.S. annually, though more probably 1.5-2 million, and no real evidence that the number is lower.
5) Kleck and Gertz themselves caution against any verification of their data outside of the amazingly high DGU number. (In a response riddled with guilt-by-association claims typical of the pro-gun agenda).
You're going to have to explain to me how you get from Kleck & Gertz cautioning against accepting one
single piece of data derived from an extremely small subset of their sample (namely those respondents who reported shooting at and hitting an assailant) to discarding
all their findings but one. See my points regarding sample sizes and confidence intervals above.
"A response riddled with guilt-by-association claims typical of the pro-gun agenda"? Care to present any arguments and/or evidence in support of that assertion? Or are you just going to cast aspersions and try to squirm your way out of it when challenged, as you usually do?
6) DGU respondents clearly had no idea of their marksmanship or other circumstances of their stories, but their 5-year-old recollections of gun use should be relied upon anyway.
Since in something in the order of 92% of reported DGUs, the respondent did not fire
at the assailant (indeed, overwhelmingly did not fire the weapon at all), marksmanship is largely irrelevant. The "other circumstances" part you've inserted yourself, conjured out of nothing except your own wishful thinking.
Contrary to what some of their critics have claimed (generally without having read their paper), Kleck & Gertz did not unquestioningly accept every claim of a DGU as valid. A respondent claiming a DGU had to provide internally consistent answers to up to 19 questions, and the incident had to meet certain criteria--unknown to the respondent--to qualify.
In one critique by Gavin de Becker (whose work I generally admire) of the Kleck & Gertz study, he cites two examples of responses given to K&G's survey-takers. In one, the respondent is pulled up at a stop sign, when a pedestrian walks a little too close to his car; the respondent draws a weapon and points it at the pedestrian, who flees. In another, a homeowner hears a noise at his living room window during the night; he shouts "I've got a gun" and the noise stops. De Becker disingenuously implies that these responses are somehow representative of the incidents counted by Kleck & Gertz, whereas in fact, these were examples of claimed DGUs
that were discarded because they did not meet Kleck & Gertz's criteria for a valid DGU: in the first example, the respondent failed to indicate that the pedestrian intended him harm, and in the second example, the respondent did not actually see anybody.
It's also worth emphasizing that these were "cold calls"; the respondents were not expecting the interview, and to generate a false positive would have had to concoct an internally consistent account of a DGU on the fly, and remember it in order to repeat it during a subsequent call-back. By contrast, as Kleck and Gertz have pointed out in rebuttals to critiques of their study, to generate a false negative, the respondent would only have had to utter the word "no" in response to whether (s)he had used a firearm in self-defense. Note also that Kleck & Gertz adjusted the numbers to give less weight to responses from the southern and western states, where firearm ownership is more prevalent than in the more populous north-east and California.
See, if you'd actually bothered to acquaint yourself with the details of the Kleck-Gertz study, rather than dismissing it out of hand because it contradicts your preconceived notions and concocting
ad hoc rationalizations, based on third-hand (at best) accounts of flawed critiques, for your willful ignorance, you might have come up with arguments that weren't such obvious chickenshit (it's too petty and small-minded to be bullshit).
7) Unlike all other message boards, comments here must be backed by legal action if they are to be taken seriously.
You didn't just make a comment, you directly accused Kleck of perpetrating scientific fraud. That's a pretty serious allegation to make about an academic. I think it'd be interesting for the boys and girls at home to see if you're willing to substantiate that assertion, or whether you're just casting mendacious aspersions in the hopes that someone will believe them but retreat under your rock when called out. Evidently, the latter is the case.
Printer-friendly format
Email this thread to a friend
Bookmark this thread
(1000+ posts)
Send PM |
Profile |
Ignore
