Science III

While I have been defending the honor of the scientific enterprise, don't think for a nanosecond that I'm uncritical. I did feel the need to point out that the descriptions of certain funded studies by a right wing web site are ridiculous caricatures, or really just lies. A complicated study of the hunting behavior and associated metabolic burden of panthers is turned into "The National Science Foundation received $856,000 in funding from the Federal Government to conduct a study to see if it is possible to train Lions to walk on a treadmill. It took the Lions eight months."  That description is a lie. You must understand that all right wing media consists mostly of disinformation. 

Now, once you know what the study was actually investigated and what methods it used, you may or may not feel that particular study is the best possible use of taxpayer dollars. That's a highly subjective question. As I have said, there's no formula that can tell us what investigations are most worth doing. We don't know where basic research will lead in the long run. You can't put every grant proposal up for a national referendum, so there has to be a mechanism to evaluate them that relies on a limited number of experts. Review panels do have specific criteria they are supposed to use, and each of the National Institutes of Health has what's called a National Advisory Council that has the final sign-off on awards. It includes lay advocates as well as scientific experts. The broad funding priorities for NIH, NSF and the other agencies that fund science are set by congress, and you get to vote for your representative and senators. I don't know what more you could ask for.

That said, science is certainly not a self-contained, self-justifying, politically and morally pure enterprise. It is done by humans, and embedded in society. It is subject to fraud, error, and various kinds of biases. Outright fraud, fortunately, seems to be uncommon, and it is fairly well-policed. If you're interested in the policing process, check out Retraction Watch.) Every human enterprise is subject to fraud, the business corporation probably more than any other. (Viz. Wells Fargo, Worldcom, and Enron.) 

Outright fraud, however, sort of blurs into error, given the nature of the enterprise. Investigators often -- actually most of the time -- want to get certain results. Whether it's because they are defending their own theories, certain dramatic findings will boost their careers, a positive finding can be the basis for the next grant, or just because positive findings are more likely to be published, they have a favored conclusion. And the rules of inference are complicated, and rather easy to manipulate, consciously or unconsciously. This is a complicated subject and I won't take time to go into it in depth right now, but the replication crisis in social psychology is a good case study. But the good thing about science is that in the long run, it's self-repairing. Spurious beliefs can hang around for a long time -- the opioid epidemic is a good example of the harm they can do -- but we catch up to them eventually.

Other biases, however, are hard to correct. The choice of what questions to ask, even if entrusted to initiating investigators, congress, expert review committees, and National Advisory Councils, is still subject to all of the cultural and power biases in society. Whether we are asking the questions that matter most to people with least power has a pretty obvious answer: probably not so much. Furthermore, the implications of scientific findings depend on your values, interests and prior beliefs. Given a finding about the public health harm done by a pollutant, what standards exactly should we set to regulate it? Given the tradeoffs among effectiveness, safety and cost of a given drug, should it be approved? To whom should it be prescribed? And racism, sexism and other biases often lead entire programs of research down rabbit holes. 

So by all means be critical. But get your facts straight.