For more than a decade, I told incoming members of the D.C. Bar as part of their mandatory ethics training that such sessions as mine were essential to making their ethics alarms ring. To support that thesis, I related the finding of research performed by behavioral economist Dan Ariely when he was at M.I.T. Ariely created an experiment that was the most publicized part of his best-selling book “Predictably Irrational,” giving Harvard Business School students a test that had an obvious way to cheat built into it and offering a cash reward for the students who got the highest scores. He tracked how many students, with that incentive to be unethical, cheated. He also varied the experiment by asking some students to do simple tasks before they took the test: name five baseball teams, or state capitals, or U.S. Presidents.
None of these pre-test questions had any effect on the students’ likelihood of cheating, except for one question, which had a dramatic effect. He discovered that students who were asked to recite a few of the Ten Commandments, unlike any of the other groups, never cheated at all. Never. None of them. Ariely told an NPR interviewer that he had periodically repeated the experiment elsewhere, with the same results. No individual who was asked to search his memory for a few of the Ten Commandments has ever cheated on Ariely’s test, though the percentage of cheaters among the rest of the testees is consistently in double figures. This result has held true, he said, regardless of the individual’s faith, ethnic background, or even whether they could name one Commandment correctly.
The classic moral rules, he concluded, reminded the students to consider right and wrong. It wasn’t the content of the Commandments that affected them, but what they represent: being good, or one culture’s formula for doing good. The phenomenon is called priming, and Ariely’s research eventually made me decide to start “The Ethics Scoreboard” and later this ethics blog.
Priming is a superb way to make sure your ethics alarms are turned on and in working order. All of us go through life focused on what ethicist call “non-ethical considerations,” the human motivations, emotions, needs and desires that drive us in everything they do—love, lust, greed, ambition, fear, ego, anger, passion…wanting that promotion, the new car, the compliment, fame, power. Good people do bad things because at the moment they are unethical, they aren’t thinking about ethics. If they were, they wouldn’t engage in the misconduct, because they would be “primed” and their ethics alarms would sound in time to stop them.
I still believe that the priming theory is sound, but it looks like Ariely’s alleged proof of the phenomenon can’t be trusted, because he can’t be trusted. Last week, an in-depth statistical analysis showed that a data set from one of the professor’s 2012 papers was fraudulent. That study had apparently demonstrated that people were more honest about how much mileage on their car if they had to sign a statement pledging that the number was accurate before they reported the mileage, rather than signing at the bottom of the page. Oops! No such study had ever been done,, and the “data” was produced using a random number generator.
Ariely responded to the claims in “humina humina” style. He said that an auto insurance company had collected the data, so it wasn’t his fault, though it was convenient that the faked results in the study lined up perfectly with Ariely’s theory. Since he didn’t check the data, the professor’s defense was that he was just incompetent and negligent, but not dishonest.
Oh. I don’t care. His employers, Duke University won’t reveal any of the details of the investigation they claim to have made into the matter. The study, which has been cited over 400 times by other scientists — is to be retracted.
I should have been suspicious of Ariely’s research earlier. Here, for example, I caught him in pushing a dubious proposition on NPR. In 2010 he told an interviewer a “fact” about the extent to which dentists agree on whether a tooth has a cavity (he said it was only 50% of the time). His stated source, Delta Dental insurance, denied his claims. And a few months ago, an Ariely paper from 2004 was branded with a special editorial “expression of concern” because there were over a dozen statistical impossibilities in the reported numbers. These couldn’t be checked, Ariely said, because he’d lost the original data file.
I have had similar disappointments with behavioral science studies and their researchers. I used social scientist Phillip Zimbardo’s 20 rules for resisting unethical influences in organizations and groups in my seminars for nearly a decade, and blogged about them here. Again, as with some of Areily’s conclusions, I have found them useful teaching tools. However, Zimbardo himself was exposed as being unethical himself, specifically regarding his Stanford Prison Experiment, which made him a national figure and launched his career in the field of analyzing culture-based misconduct. As detailed in a Medium article, his experiment was manipulated research designed by Zimbardo to reach a desired result: a lie. Worse, he has been covering up the lie for decades while still riding the wave of the fame it provided him.
When people tell us we should “follow the science” blindly because scientists are experts, those people are showing us that they…
1. ….don’t understand science, which is supposed to about nullius in verba, or “take nobody’s word for it”. Everything is supposed to be readily verifiable. Theories that are based on unverifiable claims are not trustworthy no matter who is peddling them or
2…they don’t understand scientists. Scientists are human beings, and no less corruptible, subject to bias, and governed by personal agendas and non-ethical considerations than the rest of us. Or,
3. …they are deliberately citing scientific theories and pronouncements knowing that they are not as definitive as they claim.