Updated: September 15, 2023, at 5:04 p.m.
Nearly 11 years after publishing an influential study on reducing fraud, Harvard Business School professor Francesca Gino learned her tenure would be reviewed over suspicions of the very behavior she researched.
In a series of four blog posts in June, Data Colada — a data investigation blog run by business school professors Uri Simonsohn, Leif D. Nelson, and Joseph P. Simmons — gave a detailed account of alleged research misconduct by Gino across four academic papers. One post detailed alleged research misconduct in Gino’s 2012 study on reducing fraud by asking participants to sign documents before providing their information.
Two of the papers were retracted following the allegations by Data Colada, while another had previously been retracted in September 2021. Earlier this month, a paper by Gino in the Journal of Personality and Social Psychology was retracted “at the request of the Research Integrity Office at Harvard Business School” — her fourth.
After Data Colada notified HBS in 2021 of Gino’s alleged misconduct, HBS Dean Srikant M. Datar initiated an 18-month investigation and placed Gino on unpaid administrative leave. She was also barred from campus and stripped of her endowed faculty title as the Tandon Family Professor of Business Administration.
On July 28, Harvard’s Office of the President notified Gino that her tenure was under review for revocation.
Just five days later, Gino took legal action against Data Colada and Harvard. In a 100-page filing, Gino alleged that Harvard and Datar “conspired” with Data Colada to damage her reputation with false accusations.
In the wake of the lawsuit, supporters of Data Colada organized a GoFundMe to finance their potential legal defense, which has raised more than $333,000 to date.
Andrew T. Miltenberg, an attorney for Gino, wrote in a statement to The Crimson in August that Harvard’s “complete and utter disregard for evidence, due process and confidentiality should frighten all academic researchers.”
“The University’s lack of integrity in its review process stripped Prof. Gino of her rights, career and reputation — and failed miserably with respect to gender equity,” Miltenberg wrote. “The bias and uneven application of oversight in this case is appalling. Data Colada’s vicious take-down of Prof Gino is also completely baseless.”
Miltenberg declined to comment further for this article. Gino declined multiple requests for an interview.
University spokesperson Jonathan L. Swain declined to comment for this article. Data Colada also declined comment.
At Harvard, these past months have brought more than just legal headaches. As Gino continues to battle for vindication in the public eye, faculty within HBS have grown increasingly wary of the sanctions against Gino — and the administration that imposed them.
In an internal memo obtained by The Crimson, Gino denied claims in Data Colada’s first blog post that her 2012 study on reversing the order of signing documents for honest governmental and business reporting contained manipulated data.
In the document, Gino maintained that the data fraud allegations against her were false and claimed that the findings of her study would hold even excluding the data Data Colada had flagged as having been manipulated.
“The only reason anyone would ever manipulate data would presumably be to drive results,” she wrote. “But this analysis demonstrates that even if you were to throw out all of the observations that, according to Data Colada’s own rules, should be considered ‘suspicious,’ the findings of the original study still hold.”
“Why on earth would I manipulate data, if not to change the results of a study?” she wrote.
In the memo, Gino also alleged that “Data Colada cherry-picked the data it chose to include in its analysis.”
Gino acknowledged the presence of “rows with duplicate IDs, as well as rows with out-of-order sequences” — which Data Colada claim indicate data manipulation by Gino — but asserted that Data Colada used “observations that supported their claim that I manipulated data, while omitting observations that weakened their claim.”
In their post, Data Colada pointed to adjacent rows in Gino’s data that contained duplicate IDs or out-of-sequence entries. The data show signs of manipulation, they claim, because it is nearly perfectly sorted by ID save for a few specific entries.
But Gino wrote that the blog post only mentioned two adjacent rows that contained duplicate IDs while excluding other ID duplication errors that were separated by more entries.
“When it comes to duplicate IDs, Data Colada included rows 52 and 53 (ID 49), but they omitted rows 5 and 75, which also have duplicate IDs (ID 13),” the letter read.
Data Colada declined to comment on Gino’s memo rebutting their initial post.
Though Gino denies any research misconduct, she remains unable to explain exactly how the discrepancies in her data occurred.
“I know that this still begs the question: So what actually happened in this study?” she wrote. “The honest answer is I still do not know for sure.”
Still, Gino pointed to the analog format of the study as a possible source of error.
“The original data from the study no longer exists, as the study was conducted on paper,” she wrote.
“So you can see how mistakes could creep in,” she added. “In this case, we’re talking about two duplicate ID values — sloppy and unfortunate, yes, but perhaps not terribly surprising given the paper-based system that was used.”
These errors, Gino wrote, could have been introduced by the lab manager and research assistants who performed the data collection.
“Here is what almost certainly happened: The RAs conducting the study simply stacked the paper copies by condition, and then manually entered the data in the order in which the papers were stacked,” Gino wrote.
At the time, Gino argued, there was no reason to care about the order of the data entries.
Gino concluded the letter by writing that there would be “more to come” as she continued combing through the data with her team.
“I am confident that the forensic work-in-progress we’re doing on Data Colada’s other blog posts will end up yielding strong refutations of those allegations as well,” she added.
The paper in question — a 2012 study on reducing dishonest behavior with more than 200 citations — had already been retracted in September 2021 prior to Data Colada’s blog posts this year, after the discovery of allegedly fraudulent data in a study separate from the one Gino oversaw.
In an August 2021 post, Data Colada outlined apparently falsified data present in a separate study from the 2012 paper overseen by Dan Ariely, one of Gino’s co-authors. The paper was retracted at the request of Ariely and his co-authors, who acknowledged the study contained apparently fraudulent data.
Ariely wrote in a statement to Data Colada that he did not play a role in data collection for the study and denied falsifying any data.
Gino has also disputed Data Colada’s second blog post — which alleged manipulation in her 2015 study linking authenticity to morality.
According to three professors at HBS — who spoke on the condition of anonymity, citing fear of retaliation from administrators for discussing the case — Gino believes the internal records saved by Qualtrics suggest that some external study participant had tampered with the data by filling out the form multiple times in rapid succession.
The entries that Data Colada had flagged as suspicious and manipulated, the professors said, all shared IP addresses out of China and California and were linked to old computer taglines — running Mac and Windows operating systems dating back to 2001. The emails associated with those data entries were also nonexistent addresses, two professors said.
The timing of the entries also indicated that one person was repeatedly filling out the form, as each followed one immediately after the other without any overlap — which would be expected if multiple people were responding in parallel — according to one professor.
The motivation for one person to fill out the form several times could have come from an attempt to win an Amazon gift card, an incentive offered to all who participated in the survey.
For some faculty at the Business School, the dean’s response to Data Colada’s claims against Gino revealed a wider problem.
In response to concerns over the severity of the sanctions imposed on Gino, Datar sent an email to all HBS faculty on Aug. 14 — the school’s first and only public statement since the Data Colada investigation — defending his decision.
“I did this with personal concern for Professor Gino, but also with complete confidence in the investigative work that had been done and knowing I must do what is right for our institution,” Datar wrote.
But for many HBS faculty, the email was the first time they learned of the new research misconduct policy the school had enacted and employed in Gino’s case — which she alleges was created solely for her.
According to two professors at HBS, the new policy — titled the “Interim Policy and Procedures for Responding to Allegations of Research Misconduct” — was implemented in August 2021 but not communicated to faculty until Gino’s lawsuit.
The Business School implemented the interim policy just one month after Harvard was initially contacted by Data Colada, according to Gino’s lawsuit.
Three HBS professors said the lack of consultation was an unusual move by the school’s administration, pointing to the role faculty typically play during policy creation, including small group discussions and advisory votes.
“There isn’t a single faculty member that was aware of this interim policy until Francesca’s lawsuit came out,” one professor said. “We were governed by a policy for two years that we were completely unaware of.”
The professors said some faculty are worried about Datar’s unilateral adoption of a new, school-wide policy.
The overly harsh sanctions against Gino, one professor said, resulted from “the incompetence of a dean.”
“He’s not like past deans, and he won’t be like future deans,” the professor said.
More pointedly, one professor said they believe the treatment of Gino’s case is an instance of gender inequity. Another professor pointed to examples of male faculty at the University who have been accused of — and found to have committed — research misconduct but remained on the faculty.
The professor specifically referenced Charles J. Ogletree, a former Harvard Law School professor who committed nearly verbatim plagiarism in his book “All Deliberate Speed” but remained a member of the HLS faculty until his death on Aug. 4.
Gino’s lawsuit also cited a prior investigation of alleged research misconduct by a male professor — who was not identified in the complaint — in which the school “protected the confidentiality of the male junior faculty member, and it subsequently promoted him to tenure.”
One professor said of Gino’s case that “it’s impossible to imagine that this would happen to a man.”
HBS spokesperson Brian C. Kenny wrote in an emailed statement that the school believes “Harvard ultimately will be vindicated.”
“Professor Gino has raised allegations in her lawsuit that Harvard strongly denies,” Kenny wrote. “We believe that the investigation — carried out by three senior HBS faculty members — was thorough, fair, and fully complied with HBS’s policy and procedure for dealing with allegations of research misconduct.”
“These were the first serious allegations of data falsification or fabrication Harvard Business School had received, against any member of its faculty, in many years,” he added.
Some professors at the Business School have also received emails from Harvard’s Office of the General Counsel asking them to refrain from speaking out about the Gino case or the school’s new tenure policies.
The legal hold notice — sent in August and obtained by The Crimson — requested professors “avoid any discussion of this matter outside the presence of the OGC or outside counsel.”
“This memorandum is confidential and should not be shared or discussed with anyone else without prior consent of the Office of the General Counsel,” the letter said.
The message comes in anticipation of required legal discovery of communications following Gino’s defamation lawsuit against the University.
“Any discussions about this matter, written or oral, may be subject to discovery in any subsequent legal proceeding, unless those discussions occur as part of a discussion with Harvard’s attorneys for the purpose of seeking or obtaining legal advice,” the letter continues.
The email also informed professors that they will have a “technical hold” placed on their email accounts.
“Once a technical hold is in place, documents no longer may be permanently deleted (should you delete an email, it will appear as if it were deleted, but a copy of it will be retained for litigation purposes only),” the email read.
Swain, the University spokesperson, declined to comment on the email.
In the months ahead, Gino’s defamation lawsuit is expected to move forward through discovery. Harvard and Data Colada must provide responses to Gino’s complaint by Oct. 10 and Nov. 8, respectively.
Correction: September 15, 2023
A previous version of this article incorrectly stated that Data Colada had accused Dan Ariely of falsifying data in the 2012 paper on reducing dishonest behavior. In fact, Data Colada's post outlined several possibilities for how the fraudulent data appeared, though noted that Ariely was the co-author responsible for data collection in the study in question.
Clarification: September 15, 2023
This article has been updated to clarify that the internal memo by Francesca Gino refuting the claims in Data Colada’s first blog post was not intended to be published.
—Staff writer Jennifer Y. Song can be reached at email@example.com.