Archive

Archive for February, 2010

The tragic misapplication of quantitative constructs

February 6, 2010 1 comment

As I was cleaning out the cat litter box, I overheard a statement that was, perhaps, nearly as noxious to my mind as the various scents emanating from the litter box were to my nose.  The statement was being made on a television commercial, and said something to the effect of this: “studies have shown that eating dinner together as a family will give your kids better grades and keep them off drugs.” Now, this claim should give pause to any reasonable person. What sort of study could show such a thing? How would a reasonable human being demonstrate such a proposition?

What I find most bothersome about the claim of this commercial is how it is presented as scientific. “Studies have shown…” This is scientific sounding language! But, in fact, far from being scientific, to ascribe a causal link between  “frequently eating dinner together as a family” and “getting good grades” is massively superstitious. The superstition is, of course, grounded in “correlation data” of some sort: families who eat dinner together are “more likely” to have children with good grades. But it is superstition nonetheless–it is grasping for a simple relationship between things which are obviously complicated and quite different, making connections between non-essential elements of things that you don’t really understand. This is the essence of superstition. Each category of events (“eating dinner together as a family” and “getting good grades”) admits so much internal variation that simple common sense ought to tell us that there is no meaningful way to establish any sort of cause and effect relationship between them. The well known reminder that correlation does not imply causation is a huge understatement; the two concepts (correlation and causation) are, in fact, radically different.

We ought to come to the point where we can say together: studies that show correlations are nothing more than the breeding ground for reckless speculation! These days we will let a vague and reckless usage of elementary statistical observations inform our decision making in unreasonable ways. This is, I believe, a widespread phenomena, which I would like to call “the tragic misapplication of quantitative constructs.” It is the problem of inventing measurements for things that cannot be measured, and further of subjugating ourselves, intellectually and literally, to our own measurements. It is the problem of giving to broad quantitative concepts–contrived in our own imagination–a false reality, and further of letting our measurements of this false reality play a central role in destroying the very reality that they were intended to measure. This is done through our efforts to influence or control reality by attempting to alter the measured quantity. I can think of no better source of examples of this phenomenon than those provided by the hugely successful television show The Wire.

The Wire paints a picture of the city of Baltimore from several interacting angles: throughout its five seasons, it examines the inner workings of the drug world, the police system, the school system, the political system, and the media. What we see is not good: the school system is collapsing; the police department is collapsing; the political system is wrought with corruption; the inner city is overrun by drugs and violence, the media is not telling the truth. Nobody can get anything done on any level.

As becomes evident as the show goes on, one of the major problems in the city is that each major sphere of influence has its own quantitative demands that it is enslaved to. In order to remain in power, or rise in power, the politicians need votes, and so they have to convince the citizens that they deserve them. This is done through the ability to claim certain statistics: the mayor, for instance, wants to prove that he is making the city safer, so he has to demonstrate that he has lowered the crime rate. Thus he insists that the police department deliver him lower numbers on key crime statistics (murders, robberies, etc.). This influences the decisions made by the police chiefs–often they choose to not pursue potentially fruitful investigations simply because doing so would result in “more bodies.” Or they simply lie about the stats themselves, or subtly manipulate them in a way that is detrimental to the effectiveness of the department. Likewise, the politicians want to convince the public that they are improving education, and so they must raise test scores. This standardized testing requirement renders the school itself impotent as an educational institution: they have to “teach to the test,” which results, of course, in the students not learning very much. The school and the police department both compete with each other for funding. Meanwhile, the drug dealers, whether they be low or high in the drug hierarchy, are mostly concerned about acquiring more and more money. This is done towards various ends–for low level dealers, selling drugs may just be a source of basic income. The higher ups are trying to get rich, and through that to gain things like the power or respect or control that usually come along with that.

So the whole massive web of quantitative demands becomes a cancer on the city itself. Nothing is helped, of course, by the sometimes selfish and deceitful decisions of those in power in their different spheres. Rather, the quantitative constructs that they involve themselves in encourage and exacerbate such decisions, and amplify their effects throughout the tangled and self-defeating system that has evolved and become based around them. Each one of the words in italics, in the above paragraph, was originally meant to be a measurement of something, a representation of something real. And for the person or entity concerned with that reality, changing that measurement in one way or another was supposed to indicate that a good thing had happened, for themselves at least, and sometimes for others.

The reality, of course, is that this is not so. The problem is that none of these things end up measuring what they are supposed to measure. Attempting to change them has many unexpected consequences. Seeking after votes corrupts politicians. Teaching to the test ruins the school system. Doctoring the stats ruins the police department. Getting more money does not lead to any deep or sustainable happiness for the inner city kid trying to make it in the drug game. Real people, making real decisions in their real lives, suffer hugely from the oppressive and destructive influence of the quantitative constructs that they are subject to, and subject themselves to, in various ways.

Humanity is not a science experiment. We should not subjugate ourselves to measurements that don’t actually measure anything very real. The purported concepts behind these measurements do not have the reality that they seem to have. These concepts exist, of course, for a variety of reasons, and some of them are quite reasonable, and even necessary. Money is meant to facilitate and mediate the exchange of goods and services. Voting is one of the foundations of democracy. Crime stats do, in fact, measure something about what is going on–and so do standardized tests, at least in some limited sense. But when we turn money and test scores and anything else into the goal itself, rather than a limited means to understanding something much more real, we mistake representation for reality, and thus we manufacture idols. And these idols will often war with each other at our own expense.

Categories: Uncategorized