A business school professor who I know points out that the Wall Street Journal continues to run a column by Dan Ariely, the Duke University psychology professor who has been . . . well, “disgraced” is a bit too strong, I guess; let’s just say he’s written a lot about dishonesty and then it turned out he’d published and publicized a paper that was based on fake data. Before that he published and publicized work that failed to replicate and then he responded not with serious reflection but with happy talk. And then there’s Shredders. So, whatever happened in these particular cases, this is not a guy whose reports on behavioral research are to be trusted.
To me, this is all comedy gold, and so my response to being told of the unexpected persistence of this advice column, my response was, “Dude’s got chutzpah, grant him that.”
My business school colleague responded:
Dude is one thing.
What about WSJ’s reputation? The question is related to your blogpost on the ways in which people deal with research misconduct.
Most newspapers publish multiple articles per issue. Typically they involve authors at different seniority levels and with different reputations. When a newspaper consistently publishes articles by an author known to publish and theorize about made-up facts, what does it say about other articles published in the same paper? How should it impact the paper’s bottom line?
I honestly don’t know. This column’s continued existence does seem to reflect poorly on the someone’s judgment at the newspaper, as it’s hard to imagine that nobody there has heard about the scandal of the fabricated data.
Beyond this . . . a newspaper is a big place. Several years ago I knew someone who worked at the Wall Street Journal and I asked what was the story with the credulous reporter described here. I was told that, yeah, he runs stories like that. I guess what it takes for the WSJ to keep running that Ariely column is for whoever’s in charge of that part of the budget to like the guy.
Fair enough: there’s no rule that says that someone associated with questionable research shouldn’t write an advice column. After all, Dear Abby probably didn’t know any behavioral economics at all!
This returns us to my colleague’s question about how this will affect the newspaper’s reputation and its bottom line. I guess it won’t have much if any effect. To start with, it’s just one little advice column. Also, oddly enough, it’s not clear that fraud or the association with fraud has much negative effect on institutions. I say this not from any quantitative analysis but just from considering some examples. Columbia’s most famous professor is Dr. Oz, and Columbia seems to be doing OK. Malcolm Gladwell is one of the New Yorker’s star writers. And then there’s Ted talks.
You might think of association with fraud as a bad thing, but the upside is that if you’re willing to accept a bit of fraud, you can tell better stories, and that counts for a lot! Maybe there’s a reason that the Wall Street Journal has a column by freewheeling psychology researcher Dan Ariely and not by psychology skeptics such as Greg Francis or Uri Simonsohn.
Will Dan Ariely ever get his due? He seems as fake as his p-values!
https://statmodeling.stat.columbia.edu/2008/11/02/political-attitudes-of-the-super-rich/
This has a couple of spam links in it titled “best stock trading app uk” and “private jet charter cost estimator”, have you been hacked or something?
Avi:
Thanks for pointing out. This hack happened years ago and some of these links are still around because I don’t have an easy way of finding all of them in old posts and deleting them.
Interesting discussion as always but – are you and your b-school colleague familiar with the WSJ editorial page? Dan Ariely is not their first or main credibility problem.
You mean you have a problem with the paleo-conservatives who think that child labor is just fine ?
> are you and your b-school colleague familiar with the WSJ editorial page? Dan Ariely is not their first or main credibility problem.
I think that the distinction here is: there’s a strong division of management between the WSJ’s news organization and their opinion page. The news organization has no control over the opinion page. But they probably -do- have control over Ariely’s column, which is in the “Life” section (rather than on the opinion page).
1. Dan Ariely had one case of fraudulent data used. and it’s possible he got the data from another person who made it up.
2. two more practically irrelevant stories cropped up: 2.1. his lackadaisical views on IRB, which many scientists share, as IRB are a massive hindrance to research with questionable moral justification. (see Jon Baron excellent “against bioethics” and the death in COVID by avoiding human challenge trials)
2.2. a 2004 (?) paper of his that might not replicate, or such. no one has major argument there.
3. with 400+ papers, we would’ve seem some more stuff to be found suspicious, after everyone no doubt digged into his works. I’m not aware of such.
We shouldn’t describe people based on such partial cases, IMHO
Jazi:
I dunno. I’ve published 400+ papers too, and I’ve even had to correct a few of them. But I don’t have any suspicious stories about a missing shredder etc. Nor do I respond to criticism with happy talk. This has nothing to do with IRB, and it has everything to do with published claims that I don’t trust.
I agree that this one case isn’t looking good.
he replied in detail to the allegations, and didn’t evade (I’ve read much of his replies in Hebrew. haven’t tracked the English interactions).
when it broke out, they have added two very irrelevant cases about his research. the IRB isn’t worth mentioning. the other isn’t of much note IMHO either.
the reporters however made it into a cumulative probability story
“there are three bad cases”. whereas in fact, there is only one meaningful case.
I count the lack of more cases as partial evidence of absence. given the new data investigative tools and enthusiastic investigators.
respectfully
Zazi:
There’s other iffy stuff going on; see here. Ariely was Chief Behavioral Officer at what seems to have been a fraudulent insurance company, and he’s also quoted as saying, “One of the frightening conclusions we have is that what separates honest people from not-honest people is not necessarily character, it’s opportunity . . . the surprising thing for a rational economist would be: why don’t we cheat more?”
A few other examples of questionable Ariely studies are mentioned in this news article.
Still there is something “special” about cheating in a paper about cheating. If I had made it up for a novel I likely would have been told it is too far-fetched.
The point is that “he responded not with serious reflection but with happy talk”. In other words, he missed an opportunity to address the issue, whether he was at fault or not. He could have, for example, declare how he would have done it differently by enhanced review of the data or something similar.
Compare this to Daniel Kahneman who retracted from his book the mention of several studies retrospectively found to be based on underpowered studies https://chemistry-europe.onlinelibrary.wiley.com/doi/full/10.1002/ansa.202000159.
Retracted from his book? Did Kahneman come out with a new edition explaining the errors in the original? I doubt it. It seemed to me that the book uses dubious reasoning in many places. I thought that the book would help me make better decisions, but it did not do that at all.
> What about WSJ’s reputation?
It has been a long time since I trusted anything a major newspaper wrote.
For those who missed this article, it’s a must read:
https://www.salon.com/2013/12/16/hack_list_no_10_malcolm_gladwell/
“Maybe there’s a reason that the Wall Street Journal has a column by freewheeling psychology researcher Dan Ariely and not by psychology skeptics such as Greg Francis or Uri Simonsohn.”
I was going to suggest that the latter provide some examples of advice they might give, but then I recalled from Slate that selection among advice columnists these days is to give the least helpful advice possible in hopes of going viral. Then you can also attract readers who will make up fake letters asking for advice in fictional scenarios, going ever more viral.
It’s unbelievable that anyone still takes Ariely seriously. Then again, there’s Trump.
Anon:
I don’t know about Trump, but Ariely can do amazing things with a paper shredder.
when I read Ariely’s well-known book way back when, I put it down thinking “These are some very small samples!” and that alone made me distrust the hype.
The real problem is that Psychological “Science” is a fraud for claiming association with “science”. It is not, and cannot be a science by the current definition of science (e.g., objective and mathematically formulable).
As one of a tremendous number of possible example, assume you test the memory of two groups (a control group and an experimental group. The latter is assumed to represent some notion of the workings of memory and expected to yield greater recall on average than the control).
Both groups see the same 20 words, and, after a brief delay (assumed to allow the content of short term “memory” to dissipate) are asked to recall as many of the initially presented words as possible.
After a recall interval of, say, 4 minutes, the Experimental group average 13 words and the control only 7. A statistical test shows this difference to be significant.
So, our memory hypothesis is confirmed: Our test of memorial productivity is found to be statistical reliable.
But that is All! No one can, by Any extant theory of memory, predict anything beyond “effect present/effect absent” (i.e., group A preforms better tan group B). Ordinality is the level of sophistication psychological theory (sic) can produce. No one can predict with any accuracy “how many” items will, on average be remembered as a result of theoretic consideration. 13 vs 7, 19 vs 12, 8 vs 3, etc. are all valid indices the our assumed mental mechanisms assuming these differences all attain statistical reliability. And ordinality is Not the use of mathematical tests of theory embraced by real science (of course ordinality has a place, but when it characterizes the theoretical specificity of the entire set of theories in available from a discipline, that discipline is not a source of scientific knowledge.
Sorry about the typos in the above post. I am a typo generator.
sk