rebecca
patreon
One Disgraced Scientist Gets an NBC Show, Another Becomes a Libel Bully
🕑 Added 2023-09-08 21:51:25 +0000 UTCHey patrons! Enjoy this video early and ad-free all weekend long! It'll go live Monday morning. Oh and yes I did break down and dye over the red to better match my roots. Oh yeah and here's another BlueSky code: bsky-social-63cbt-elrn2
Transcript:
I remember when I was in elementary school, maybe at age 8 or 9, I was doing a science project that, of course, required the standard “make a hypothesis, test it, collect the data, evaluate the data, publish the results” stuff. And I remember getting some really messy data, and thinking, hey, it would be super easy to just add in a few more data points and get a really clear and obvious validation of my hypothesis. And so I did, and as I was putting my little posterboard display together I remember feeling absolutely fucking horrible about it. Like, even though there was NO WAY my teacher would know, *I* knew, and Jesus knew (I believed in Jesus back then), and so I went back and put in the original data and my result was really messy and I did not win the science fair. As an adult I now know that the result didn’t really matter: the entire point was to teach me about the scientific method. No one really gave a shit which paper towel absorbed water the fastest. But I do think I learned an additional, unintended lesson: without careful oversight or reproduction, someone who is MORE motivated or LESS ethical than myself at the age of 8 can easily get away with “science” that’s based on pure fantasy.
And so, I’m once again here to talk about this guy: Dan Ariely, author of books such as Predictably Irrational. I last discussed Ariely almost exactly two years ago, when a team of independent researchers dug into the data on one of his most famous studies and found very clear, very convincing evidence that Ariely himself had made up data to get the conclusion he wanted. The ironic kicker was that the study in question was about honesty – can you “nudge” people to be more honest by asking them to sign a pledge of honesty before they self-report something they might be motivated to lie about, like a tax return? If you haven’t watched that video I do recommend you go back and give it a go because it’s a really interesting story of how even his own co-authors were duped, but later came around to realizing their own study was complete bunk. It’s also fun to learn how a simple mix-up in Ariely’s choice of font made it clear that it wasn’t just an honest mistake (as he claimed that he just used the data given to him by an insurance company), but likely out and out intentional fraud.
I’m circling back to this story because I have a few very important updates. First, because the insurance company that Ariely blamed for the fraudulent data did eventually weigh in. In July, they gave a statement to NPR “confirm(ing) that the data it had provided for that study had been altered after they gave it to Ariely, but prior to the research's publication: "It is clear the data was manipulated inappropriately and supplemented by synthesized or fabricated data." In case anyone had any remaining doubts. They even applied Ariely’s supposed statistical analysis to the actual data they say they gave him, and found that it “does not contain a statistically significant difference between those who signed forms at the beginning and those who signed forms at the end, so the hypothesis/conclusion around honesty for “Sign Beginning” and “Sign End” would not be warranted.”
But wait, there’s so much more! Here’s a second update: since all this came to light in 2021, Ariely has been fired from his position at Duke University, canceled by his book publisher, and chased out of psychology entirely to the point where he’s been forced to pay bills by joining the crew of an Alaskan crabbing vessel.
Just kidding, he’s still a professor at Duke University, his publisher Harper Collins makes no mention of the fraud on their website, and this fall NBC will be airing the hot new police procedural The Irrational, based on the book Predictably Irrational and starring Jesse L. Martin as “Alec Baker, a world-renowned professor of behavioral science, (who) lends his expertise to an array of high-stakes cases involving governments, law enforcement and corporations with his unique and unexpected approach to understanding human behavior.” Ariely, of course, will be serving as consultant.
And according to IMDB, the hero will be facing the most horrific of nemeses: someone whose “behavior is unable to predict.” So like…a cat? Murdock from the A-Team? A Massachusetts freeway driver?
So yeah, this alleged fraud has actually been rewarded with his very own TV show that will only boost the myth that his research is legit. Cool, great.
And wait, there is yet more. I should mention that the reason why I was looking for updates on Ariely is because one of my patrons, Chris, asked about it in one of my recent monthly livestream Q&As. Chris linked to an article that described accusations against both Ariely AND Harvard Business School professor Francesca Gino. Gino was a coauthor on that “sign at the top” honesty paper, but the original Data Colada reporting really seemed to point to Ariely as the one responsible for the data manipulation. But I had missed that in June of this year, Data Colada’s researchers published a four-part series showing extremely convincing evidence that Gino had ALSO falsified data IN A DIFFERENT STUDY IN THAT SAME PAPER. “Signing at the beginning makes ethics salient and decreases dishonest self-reports in comparison to signing at the end” involved three studies, and at least TWO OF THEM were DISHONEST.
As always, links to all sources can be found in this video’s transcript, linked in the dooblydo below, and I encourage you to read Data Colada’s report in full, but I’ll summarize what they found: in Gino’s study, the data seemed to be organized by participant ID, which is the leftmost column. But Data Colada has highlighted several times where the ID appears out of order or duplicated, leaving the data in an arrangement that is impossible for a spreadsheet program to sort, which means that someone must have sorted by ID and then gone in and moved or edited those cells.
And what are the chances? Those suspicious cells are on the extreme ends of the results, which tilts the results heavily in the direction of the prediction, which is that people are more honest when they sign at the top of the page. They further show evidence that someone literally went into the Excel sheet and dragged those lines to the opposite condition to achieve the desired result.
That’s all bad enough, but Data Colada goes on to detail THREE MORE studies that Gino authored that have evidence of data tampering, and they write that they suspect there could be DOZENS more. Their posts are based on a full report of their findings that they turned over to Gino’s employer, Harvard Business School, in the fall of 2021. They did their own review, resulting in an approximately 1,200-page internal report, suggesting that they not only confirmed Data Colada’s findings but possibly found a LOT more to be upset about.
Unlike the situation with Ariely, Gino did not end up being rewarded with a television show. The studies Data Colada exposed have been retracted, and Harvard updated their ethics policy to make termination a possible punishment for research misconduct. According to Gino, Harvard Business School’s dean told her “that she would be put on unpaid administrative leave, immediately revoked her named professorship, and told Gino that he requested the University president…that Harvard (begin) proceedings to revoke Gino’s tenure. Datar also informed Gino that she would be barred from campus.” If successful, this would be the first time Harvard has ever revoked tenure from someone.
So there we go, finally a happy ending, right? Well, not exactly. I know all this, about Harvard’s updated policy and Gino’s supposed punishment, because Gino filed a defamation lawsuit against Harvard, the dean of Harvard Business School, and–no surprise here–the three analysts behind Data Colada.
That’s right, it’s yet another case of an ALLEGED conniving, lying, immoral pseudoscientist getting caught and responding with a defamation lawsuit, which IN MY OPINION seems to be simple libel bullying, done in a desperate attempt to stop criticism.
In descriptions of the lawsuit, I can’t find anywhere where Gino disputes Data Colada’s findings that the data in her studies was fraudulent. Instead, she seems intent on arguing that they can’t prove that SHE is the one who manipulated the data in the four studies that she oversaw with different co-authors on each one. I mean yeah, maybe the field of psychology is so absolutely fucked that you just happened to publish four different studies in which four different scientists secretly manipulated data and you didn’t notice. I’m only being a little sarcastic there, MAYBE THE FIELD OF PSYCHOLOGY IS ACTUALLY THAT FUCKED.
Gino is also accusing Harvard of sexism, which is pretty hilarious because on the one hand it is absolutely absurd to ALLEGEDLY get caught red-handed doing something wrong and then arguing that you’re only getting punished because you’re a woman. But on the other hand maybe she has a point, because again, Dan Ariely got a TV SHOW. But yeah it would be like a child predator arguing that it’s not fair you’re sending her to prison while Roman Polanski wins the Grand Jury Prize at the Venice Film Festival. You’re right, it’s not fair, but not for the reason you’re implying.
Anyway, she’s probably going to lose her stupid case against Data Colada but that’s probably not the point – the point IN MY OPINION is to drag them through legal hell for a few years as punishment for catching her. They’ve set up a GoFundMe to fight back but the good news is that as of this recording it’s at a very healthy $330,000 AND they have high hopes that Harvard will step up to cover the legal costs, granting them the freedom to “pursue what's right instead of what’s most affordable.” If only our most famous psychologists had the same sort of ethical mindset, maybe we wouldn’t be talking about this at all. You know, I wonder if it would have made a difference if scientific journals had researchers sign an honesty pledge at the top of their study submission form? Just an idea, something to think about.
Comments
Robert Keefer
Proud to say, one of the authors of Data Colada, Joe Simmons, was my student as an undergraduate; as close to fame as I'm ever likely to get. He and the other two wrote was is (I believe) the most cited article in the history of the lead journal of the APS (Associate for Psychological Science), Psychological Science in 2011, which was also about keeping science (specifically psychology) honest.
But should they sign the honesty pledge on their research before they submit it or after it is published? What does the research show?
