Who knew there was a code of omerta in academic research? We’re told that academic research is pure science, above reproach, perhaps debatable, but never dishonest.
But the saga of Michael LaCour and the young researcher who doggedly exposed him shows there is dishonesty in academic research and an atmosphere of fear among those who would tell the truth.
Graduate student Michael LaCour, along with noted Columbia professor Donald Green, published a paper in the journal Science that purported to show interviews conducted by self-identified gays with opponents of gay marriage would have the long-lasting effect of changing the person’s mind in favor of gay marriage. The simple act of a self-identified gay man talking to a gay marriage opponent would change minds more or less permanently.
Such a proposition was explosive and was covered in the academic and mainstream press all over the country. It has since been thoroughly debunked. LaCour has been charged with being a serial liar and a liar even on matters not that important. For instance, even though he had won many awards in his work, he felt it necessary to lie on his CV about receiving an award that does not exist.
The more serious lies involve the research itself. The data set he claimed was his own was really from others. What’s more, the company he claimed to have used to develop the date set had never heard of him, and couldn’t have done such a complicated research project anyway. Finally, the data set does not show what he says it shows.
LaCour’s trail of dishonesty is a fascinating look into how he fooled major researchers and an academic journal, mostly through his own winning personality, obvious talent and being protected by coupling his name with a prestigious Columbia University professor.
Jesse Singal at the Science of Us site at New York Magazine tells this detective story in great detail.
Third-year doctoral student David Broockman was blown away by early results of the LaCour study. He met LaCour for breakfast during the 2013 Chicago meeting of the American Political Science Association who showed him yet more results on his iPad. According to Singal, the sexy thing about the LaCour results is they flew in face of most existing research that shows opinions on such controversial issues tend to stay fixed and would never change and stay changed after a brief conversation.
Broockman was so excited about the research that he wanted to do his own along the same lines, replication being something quite common in academic research. Immediately, though, something didn’t quite add up. The cost for what LaCour said he had done — face to face interviews with 10,000 people — would have been at least $1 million, an amount of money no graduate student would be able to be get his hands on.
Broockman called the company LaCour said he used for canvassing and the salesman said he was not sure it could do such a large job. Not wanting to rock the boat, Broockman didn’t ask why they could do it for LaCour and not for him. If he pushed, he would have discovered LaCour had lied about any relationship with the company and that the company had never heard of LaCour.
Instead, Broockman wrote to a number of other polling firms describing what he wanted to do and none of them said they could do it. Broockman says at that point he began to wonder if the LaCour study was a fake.
Broockman eventually partnered with another graduate student, Josh Kalla of UC Berkley, and together their suspicions continued to grew, particularly as they began to pick apart the data which at that point had been published by Science.
Their suspicions about anomalies in the LaCour data are very technical and of interest only to academics in this field. According to Singal, “The data was, in short, too orderly given that it came from a big survey sample.” Singal says Broockman’s excitement-to-suspicion ratio went from 90-10 to 50-50.
Here is where omerta comes in, an academic code of silence punishable by academic death. Repeatedly Broockman was told to back off. Professor Neil Malhotra of the Stanford Business School, told him: “As someone in your early career stage, you don’t want to do this. You don’t want to go public with this. Even if you have uncontroversial proof, you still shouldn’t do it.” Questioning someone else’s work could be death and questioning the work of someone like Professor Donald Green of Columbia could be certain death.
Indeed, LaCour’s selection of Green as a collaborator on his phony project was perfect. So well respected is Green that others who had deep suspicions about the study — including Stanford’s Jon Krosnick — accepted it when they discovered Green was a collaborator.
So, Broockman took the unusual step of putting his suspicions anonymously on a site called poliscirumors.com, what some refer to as a “cesspool” of academic rumor and backbiting. Even there, his accusations were controversial; the anonymous site host took them down.
In the meantime Broockman had been hired to do a transgender study in Miami [one thing you note in this story is the huge money that’s available for studying anything LGBT] and he became discouraged by the response rates he was getting. Less than 1 percent of 13,000 mailed questionnaires were returned, whereas LaCour had claimed a response rate of 12 percent on this first try. Broockman asked a LaCour research partner just how LaCour got such a high response. The result was an email message through LaCour from someone at uSamp named Jason Peterson. And here is where things got very strange. A call to uSamp revealed Jason Peterson did not exist. The firm also confirmed it couldn’t even do the research as described in the LaCour study.
Even then, Broockman was warned off. Don’t rock the boat. You’ll get hurt.
Eventually Broockman and his colleagues discovered the data set that LaCour cribbed from, the one he not only passed off as his own but manipulated to show results that were phony and that the media lapped up hungrily.
It was the final smoking gun. Last week Broockman, and his fellow academic sleuths — UC Berkley grad student Josh Kalla and Yale political scientist Peter Aronow — published their explosive findings that the LaCour study was totally false.
Even now Princeton University has not yet withdrawn its offer for Michael LaCour to join the faculty next month. For his part, LaCour continues to deny the charges and has issued a response he says rebuts the Broockman report.