Lysenkoism Watch: The "Data Quality" Act

Check out today's WaPo article on the "Data Quality" Act. This is the first I've read about the act, so of course I can't speak with any authority about it, but I can say some general things about why we should be concerned enough to pay close attention to this legislation.

As you know and as the article makes clear, many (perhaps even most) arguments in science--and out of formalized science, for that matter-- involve assessing the weight of conflicting evidence for and against some hypothesis. We can think of this as a kind of UNbalancing, analogous to weighing two items of extremely similar mass to determine which is heavier. If we build an inaccurate scale--or if we put a finger on the scale, continually nudging one side up or down depending on some antecedent preference about which we'd like to weigh more--the results of our efforts are basically valueless.

One way of cheating the system is to employ illegitimately differential standards of proof. Now, different standards of proof are appropriate in relevantly different cases. If John merely dislikes peanuts but Jane has a deadly allergic reaction to them, then it is rational for Jane to demand more proof than John that the dish before them contains no peanuts. But, if the WaPo article is getting it right, this is not the purpose of the Data Quality Act, nor is it how it is being used. According to this article, the DQA is being used to nudge the scales in favor of big business. When the weight of evidence indicates, for example, that some substance is harmful, this act allows those who oppose regulation to arbitrarily raise the required standards of proof.

Given that evidence is often available on both sides of any important question, if you give me the ability to raise and lower the standards of evidence at my whim, I can frequently torture the data to support whichever conclusion I prefer. The application of systematically differential standards is an old trick of pseudoscientists. A good example here is "Creation Science," which basically demands that evolutionary theorists adduce irrefutable evidence in support of their hypotheses, while rejecting these same hypotheses on the basis of even the flimsiest of objections. Creationists, of course, treat their own claims in exactly the opposite way, accepting even the most preposterous arguments in favor of their claims, but refusing to accept even the weightiest evidence against them. Creation Science is an extreme example of the illegitimate use of differential standards of proof, but a far slighter variation in standards of proof will wreak similar havoc when the weight of confirming and disconfirming evidence is more equally balanced.

The (ab)use of the DQA reported by the Washington Post is, of course, consistent with the attitude of the Bush administration to science and to rational inquiry in general, including the collection of intelligence data leading to the WMD debacle in Iraq. The WaPo article indicates that, as many of us have been saying, Lysenkoism is alive and well in the Bush administration, the Postmodern presidency in which conclusions are not accepted on the basis of evidence, but, rather, evidence is accepted or rejected depending on whether it fits with antecedently-held, ideologically-motivated conclusions.