Pages

Saturday, November 11, 2023

The latest on the Harvard data manipulation affair

The Harvard Crimson reports that co-authors of suspended professor Gino are looking to distance themselves from the fallout:

Harvard Business School professor Francesca Gino, an academic known for her studies on dishonesty, is at risk of losing her position at the University over allegations of data fraud. Now, Gino’s former co-authors are working to keep the rest of her research honest. 

This Monday, six of Gino’s co-authors — professors Max H. Bazerman, Julia A. Minson ’99, Don Moore, Juliana Schroeder, Maurice Schweitzer, and Uri Simonsohn — released the “Many Co-Authors Project,” a compilation of raw data from all of Gino’s past studies. By providing an avenue for academics to review Gino’s work, they hope to correct the scientific record and protect Gino’s co-authors from allegations of misconduct. “We see it as our professional and moral obligation to provide information about the data provenance and data custody for those papers,” the project’s mission states.

This past spring, the data analysis blog Data Colada alleged there was fraudulent data in four of Gino’s studies. Simonsohn, one of the organizers behind the Many Co-Authors Project, is also a writer for Data Colada. The site’s findings prompted an HBS investigation into Gino’s work, which culminated in Gino being placed on unpaid administrative leave earlier this year.

After she was placed on leave, Gino leveled a $25 million lawsuit against the Business School, Data Colada, and HBS Dean Srikant M. Datar. Gino maintained that she was innocent and accused the defendants of conspiring to damage her reputation, alleging gender discrimination in the school’s handling of her case. In a public statement following the release of the Many Co-Authors Project, Gino’s team voiced concerns that the project could interfere with the lawsuit. “Instead of enhancing research practices of the field at large, the project is a witch hunt for one woman, instigated by the very people and groups involved in the lawsuit,” said Andrew T. Miltenberg, an attorney for Gino.

The six researchers emailed 143 academics across 138 published papers co-authored by Gino, asking whether Gino had been involved in the data collection of each paper. If the co-authors confirmed or were unsure of Gino’s involvement, the Many Co-Authors team requested access to the papers’ raw data. The project’s website catalogs Gino’s involvement in each of these studies and links the raw data when available. The project also emailed Gino a spreadsheet with questions on the 138 publications, which they said they would publish if Gino provides a response. In the spreadsheet, the six researchers asked about Gino’s involvement in data collection, whether the data was collected on Qualtrics, and for collaborators with authorization to Qualtrics for each study.

Gino criticized the Many Co-Authors Project for a lack of transparency, claiming the group kept her “in the dark” and only invited her to participate two weeks before it went public. “I have deep reservations with the way the project was structured,” Gino wrote. “For months, I was kept in the dark about the details of the MCAP.”

According to the website, the Many Co-Authors Project will make all information it receives publicly available, and studies’ authors are encouraged to caption their articles with updates on the presence or absence of apparent data manipulation. These “living documents” can be changed as academics reexamine the validity of the studies they co-authored with Gino...

Full story at https://www.thecrimson.com/article/2023/11/8/gino-many-coauthors-project/.

(If you haven't been following this matter, type "Gino" in the search engine for this blog.)

No comments: