Please don't sue the scientists who exposed your papers for containing bad data

"Data analysis just got dangerous"

Art for The Intrinsic Perspective is by Alexander Naughton

A trend over the last few weeks is that some incredibly high-profile scientist: Dan Ariely at Duke, Francesca Gino at Harvard, the literal president of Stanford, Marc Tessier-Lavigne, is accused of data fabrication/manipulation and scientific misconduct more broadly, and then dragged over the coals in the press and ejected from or demoted by their institution.

Certainly, the subject matter of some of these scandals, like fabricated research on honesty, itself helped attract press attention.

NPR

In many ways, I’m sympathetic to those accused under such an intense limelight. I don’t think findings of data fabrication or data mangling in a small number of papers out of hundreds of published papers universally merits career loss. Not automatically at least. I do think intent matters for the harshest of judgements, and in all cases the authors themselves deny knowing about the problematic data in their papers. Additionally, I’m not one for public condemnation in general.

But yesterday it was announced that Francesca Gino, one of the accused researchers, is now suing the very scientists who exposed the likely data manipulation in several of her papers on their joint blog, as well as Harvard University for firing her, in a 25 million dollar lawsuit. According to The Chronicle of Higher Education:

In her lawsuit, Gino said that she had been placed on a two-year administrative leave on June 13, which bars her from campus and strips her of her salary, teaching duties, research responsibilities, student mentorships, and titled professorship. She also alleged that the allegations have “sullied if not destroyed” her career, and have delayed a book deal.

If 25 million seems a lot keep in mind—these aren’t just professors. A lot of her research is implicitly geared toward businesses, like how to get people to be honest on forms they fill out, etc. And a Harvard Business School Professor is more like a scientist-celebrity, as attested to by Gino’s speaking fees.

source

While the bloggers, a group called Data Colada, were the ones who first brought the issue to Harvard’s attention (without going public), a two-year internal investigation at Harvard culminated in Gino being placed on administrative leave, as well as an internal ~1,200 page report confirming the bloggers’ general conclusion the studies had problematic data (and the bloggers finally published their own conclusions afterward). Harvard even hired an independent analyst to look into the data. All four of the papers originally identified as problematic have now been retracted or are undergoing the process of being retracted. However, according to The Chronicle:

Gino also argued that Data Colada’s scrutiny of her—and Harvard’s treatment of her during the investigation—amounted to gender discrimination.

Gino’s lawsuit described her as “a working mother of four young children and the breadwinner in her family,” and someone who has “received praise by her female colleagues and collaborators for serving as a role model for other women at HBS.” She wrote that the Data Colada authors “have targeted the work of prominent female academics and subjected it to an exceptionally high level of public scrutiny.” She cited the case of Amy Cuddy, a former Harvard Business School professor whose work on “power posing” was criticized by Data Colada and who subsequently left Harvard.

The problem I foresee is that Francesca Gino’s lawsuit creates the obvious following conclusion: if you identify cases of what you think is scientific fraud from a prominent researcher, you might now be on the hook for millions, unless you, in your accusations, can directly prove intent with a smoking gun (rather than merely that the author has papers where the data is highly problematic). The implications are vast.