Shannon Brownlee, a senior fellow at the New America Foundation has penned an interesting and in depth article about how conflicts of interest warp medical research, available here.
What’s noteworthy here, IMO, is not just the basic message that conflicts of interest can influence researchers or publishers or institutions and that, therefore, as in any other field of human endevor where this issue arises, we might want to have some systems in place that at least disclose the conflicts or, better, try to minimize them. Rather, what’s noteworthy is the belief prevalent among scientists and others that somehow science escapes this universal truth because its, well, science.
I started this a long time ago, and just dug it up again. I decided to run it rather than delete it because I worry that with the current assault on science from some quarters, it is easy to forget our own prejudices and warts. and we shouldn’t, or they will bite us in the rear as assuredly as ignoring real science does.
In any event, the Shannon Brownlee piece above highlights one of these warts. The study of science is a noble pursuit, and there are indeed facts that are incontrovertible reality. But a lot of science is interpretation of data and drawing conclusions, a slow and painful process subject to human error. As Robert Heinlien once observed, a scientist can be just as narrow minded and ignorant on subjects outside his (or her) specialty as anyone else.
Nevertheless, the idea that science, because it is the pursuit of facts that explain the universe, renders its practitioners immune to human failings and prejudices, has numerous adherents. This attitude impacts not just conflicts of interest. Scientists and engineers have vigorously resisted the idea that somehow the diversity argument applies to them, because science is science. Therefore, some argue, efforts to actively recruit underrepresented groups or criticism of studies based on lack of inclusiveness, represent the worst sort of “political correctness” and constitutes a threat to real research.
Bunk. Turns out that scientists and engineers are human beings like everyone else, and we should start considering the implications of that statement.
One of the implications is that we like to surround ourselves with people like ourselves, and who basically agree with us. For many of us this is a harmless prejudice. But what happens when you are in charge of reviewing data that impacts peoples lives? When you have the power to make tenure decisions? When you have the power to distribute millions of dollars in career-making grants?
The other issue in being human is resenting when people point this flaw out. First, it sounds like a specific criticsm of the individual’s character. As Brownlee reports, many scientists she talked to interpret the statement “human beings are influenced by people who give them money, when you take drug company money it may influence your results and those evaluating your results” as “you are a bad scientist who will lie for money.”
Friends of mine in the sciences tell me this is a big problem for folks trying to have an academic career. The universe in which you live is small, it’s thinking controlled by a relatively few individuals with access to money and who review writings for publication, and determine tenure. As a consequence, it becomes difficult to provide interpretations of experimental data or design experiments that challenge accepted norms. Gallelio might no longer get tortured, but he would find it difficult to get published or secure research grants.
The danger, of course, is to go all the way and decide that there is no such thing as real fact — only prejudice and interpretation. But I’d like to suggest a different conclusion. The true value of science lies in its repeatability. If a fact is a fact is a fact, I can repeat it elsewhere.
And this is the value of diversity, particularly in the sciences. We should fund dozens of different teams to review each others work. We should want as many different folks from as many diverse skill sets as possible. Experiments produce data, but the interpretation and organization of that data is a uniquely human function. To maximize our chances of success, we must recognize the human element and embrace it, rather than deny it as some how contrary to science.