Thursday, January 14, 2016

Scientific Deception

Recently, I had a conversation with a graduate student friend of mine about how a scientific education would be perfect training for a snake oil salesman. In addition to all the experimental techniques and methods of investigation you learn in graduate school, you also learn a skill that must be vital for a bullshit artist, and that’s the ability to sound like a scientist. Scientific-sounding stuff has incredible power over the average non-scientist, since most people are willing to defer to an expert authority regarding things they don’t know much about. Even the Trump campaign, which has buried facts in a shallow grave, uses scientific-sounding facts and statistics to try to sway people. The point is that science as a discipline is in a position of intellectual authority for many people, and its very easy and potentially tempting to use that power to manipulate people for personal or ideological gain.
I think the article “Half the cancer drugs journalists called “miracles” and “cures” were not approved by the FDA” makes a good point about this. Making up data, as we’ve seen in the Trump campaign example or in the STAP fiasco in Japan, is only the most egregious example of scientific misconduct. Simply overstating a claim or using an excessively strong superlative is a form of deception. Calling a new cancer drug a “miracle” isn’t likely to cause significant harm by itself. In fact, in the scientist’s view that tiny deception might allow a paper to be published in a higher impact-factor journal or just pump up an individual’s ego. However, these kinds of overstatements, when directed to the non-scientist public, can seriously undermine the public’s trust in science as an authority.
These small deceits fit nicely with the conclusions in Dan Ariely’s TED talk. However, I’d like to point out a small example of deception in Ariely’s talk itself. I noticed that when describing his study subjects, Ariely would give some details about the study population—most of the studies he discussed in his talk used college students or university employees as subjects (Here’s one example.). However, when Ariely made his closing statements he substituted “college students” for “people”. Of course, college students are people, but switching the scope of a conclusion from college students to people in general seems like a huge jump. Would you reach the same conclusions if you changed the study subjects to Georgia State Fair attendees? What if they repeated the study in Syria? I don’t know, but it certainly seems that you’d need to sample more than 100 Carnegie Melon students to make meaningful conclusions about universal human behaviors. I’m sure Ariely didn’t intend overt deception, but it goes to illustrate just how important precise, scientific language is, especially when talking with a lay audience.

No comments:

Post a Comment