Tuesday, December 7, 2010

Social Science Palooka

David Brooks seems sneaky. He writes like a rational human being, but always throws in a "whammy" that tends to discredit the assumed thrust of his article. This is what some people call "polemics."

His facile conclusions never fail to disappoint. This December 7, 2010 column in the New York Times is no exception.

Social Science Palooza

Ask any Republican or other member of the right wing what he thinks about the social sciences. You will receive either a blank stare, or a sneer. So, when David Brooks wrote an article about the social sciences, I wondered why.

The first thing I learned is that he finds such studies "bizarrely interesting." Which indicates to me, a social scientist, that he finds it difficult to grasp the scientific methods used to conduct these studies. Results based on the Scientific Method are not bizarre.

When I review social science studies, I always verify the underlying data before I apply it to new information. It doesn't take long. Really. It's a non-starter when the person writing the article quotes himself or the organization who paid him to write the article as far as maintaining objectivity.

This observation will explain what I mean. "For an article in The Review of Economics and Statistics, Mark Duggan, Randi Hjalmarsson and Brian Jacob investigated whether gun shows increase crime rates. They identified 3,400 gun shows in Texas and California and looked at crime rates for the areas around the shows for the following month. They found no relationship between gun shows and crime in either state."

Hypothesis: Gun shows increase crime rates.
Underlying science: Economics and Statistics

The problem: 1. Statisticians always allow for a margin of error in their analysis. 2. Therefore, the statement that "no relationship between gun shows and crime" could not be based in fact. 3. Because there HAS to be a variation in the data, just because we are human beings and inconstant. Therefore, Brooks should have stated the margin of error. 4. And, crimes that occur around gun shows are not always violent, and can involve fraud or misrepresentation.

For example, one of my friends bought a gun at a gun show, to be delivered later. He paid his money, but the gun never came. It took research and a trip to the town where the gun dealer lived to involve the police for my friend to get his gun months later. That counts, eh?

So, when one of these rational-sounding people cannot come up with a rational explanation for normal phenomena, one must assume that they are just a few bricks shy of a load. Or perhaps just Social Science Palookas.

No comments: