Lotta Nonsense
ArchPundit
Get caught up on the Lott Affair over at Tim Lambert’s.
A couple interesting points, Lott has not responded to the data miscoding. This fits a pattern with him where he is questioned about something and attempts to change the subject. It happened to me in discussing his surveys. I tried to get him to answer the question of what the point of the surveys were if they had margins of error that were so large they were unuseful. Lott got angry over some side issues and never answered the question claiming he was being misrepresented.
The irony is that he serially misrepresents other’s work.
Reynolds is also playing both sides of the fence claiming he isn’t qualified to evaluate the statistical work while slagging on ‘antigun’ research methodology. I’ll agree that much of the popular stats thrown around to support gun control are deceiving, but I’m qualified to comment on all of the statistics. Of course, Tim Lambert points out that many of the criticisms of the work Reynolds seems to be reffering to are inaccurate.
Fundamentally, Reynolds misunderstands social science. He attacks two methods of proxies for determing rates of gun ownership as wanting.
What’s most striking to me, though, is another study, by antigun researchers, that tries to measure gun ownership by suicide rates. (And it’s not mentioned here, but I believe there was another that tried to use subscriptions to gun magazines as a proxy.) This seems rather bogus to me, and I can only imagine the general derision if this kind of proxy were employed by researchers whose work supported gun ownership.
As Lambert points out, Kleck uses the suicide rate in a published paper. Is a proxy perfect? No–pretty much by definition a proxy is used to represent another variable that is unavailable, but highly correlated. Some of the time we find great proxies, other times, we try proxies and they turn out not to work. Testing them for effectiveness is a part of the social scientific process. In the case of the Duggan paper, the use of gun magazine subscriptions as a proxy was a reasonable attempt. It failed and later it was found to be non-robust (IIRC). So we throw it out and move on to better proxies. That is the social scientific process.
Reynolds seems to want to see social science as a one-time shot in which all of a sudden we find the evidence in a Eureka! paper and simply accept the findings. This is not representative of social science, or for that matter the natural sciences. Findings are tenatively accepted until they are confirmed, extended or falsified. They aren’t set in stone. Some treat them that way, but the process eventually wins out.
Ridiculing an attempt that was tried and discarded is a bit strange because it demonstrates the process working. Instead of having an argument about whether a proxy would work, people tested it. Maybe this is an extension of the problem Reynolds has understanding the difference between advocacy in the legal academia versus testing hypotheses in social science. Ultimately, whether a proxy is good or not is an empirical question. Reynolds seems to view it as common sense when few people actually have the knowledge to understand what common sense would determine to be a good proxy if they do not have a fairly strong understanding of statistics. Despite his self-professed lack of understanding of statistical analysis, Reynolds makes strong claims about the accuracy of specific proxies without understanding what the process of how a proxy came to be.
One of the jokes about a prominent political scientist, William Riker, is that one knows he is important because every major finding of his was disproved. There is not a single magic bullet paper in social science, though there are papers that set the research agenda. Assuming that any single paper settles the questions as Reynolds wishes Lott’s research to do is a terrible misunderstanding of science. And it points to the ultimate problem with Lott–he doesn’t think he could be wrong. Instead of taking the criticisms of his data and methods and working to make them stronger based on those criticisms, Lott declares victory unilaterally.