A McMaster University scientist and holder of a prestigious Canada 150 research chair is facing scrutiny after two peer-reviewed studies were retracted and collaborators made public statements that data he provided to them contain duplications and other irregularities.
Retractions have been requested or are pending for five additional studies, while data underlying nine others have been flagged for unusual anomalies, according to a spreadsheet one former co-author is keeping.
The unfolding controversy has rocked the field of behavioural ecology, in which the scientist, Jonathan Pruitt, an associate professor at McMaster since 2018, is regarded as a star performer.
A spokesperson for the Hamilton university confirmed that its office of academic integrity is looking into the matter. McMaster’s research integrity policy requires the office to initiate an inquiry to determine when allegations of academic misconduct merit an investigation.
Dr. Pruitt is conducting field research in the South Pacific and did not respond to e-mails from The Globe and Mail requesting comment. A story published on Friday in the journal Science said that, when contacted by a reporter last week, Dr. Pruitt denied fabricating his data, but added that errors can happen in data management.
Dr. Pruitt’s research uses spiders to test ideas about the evolution of behavioural differences among individuals in response to environmental and social pressures. The work had gained media attention because of its relevance to understanding how communities of organisms survive.
It was through his scholarly output that Dr. Pruitt became one of 24 top academics recruited to come to Canada from institutions in the United States and elsewhere as part of the federal government’s Canada 150 celebrations. As a Canada 150 research chair, he is guaranteed support for his research program at McMaster amounting to $350,000 a year for seven years, or $2.45-million total in public funds.
Questions about Dr. Pruitt’s work came to light on Jan. 17, when a paper he co-authored in 2016 while a faculty member at the University of Pittsburgh was retracted. One of his co-authors, Kate Laskowski, an assistant professor at the University of California Davis, described in a detailed blog post how she became aware of issues with the data Dr. Pruitt provided for the study.
Since then, at least half a dozen co-authors and former students of Dr. Pruitt’s have reported similar problems.
“The response from the collective community has been swift and very transparent,” said James Heathers, a computer scientist at Northeastern University in Boston and a specialist in forensic analytics, which is concerned with error detection in research data. “I’ve never seen anything like it.”
Dr. Heathers said he has examined some of Dr. Pruitt’s raw data and confirmed that systematic anomalies affect several research publications.
“At absolute best it might be a colossal oversight,” he said, noting that poor data collection and management practices have tripped up experienced researchers in the past. However, he added that the extent and nature of the problems would also lead anyone examining the matter to “start discussing the role of intent.”
In Ottawa, Margaret Blakeney, a senior policy analyst with the federal Secretariat on the Responsible Conduct of Research, would not comment directly on Dr. Pruitt’s case, but said every holder of a Canada 150 research chair must adhere to integrity guidelines “as a condition of applying for or holding agency funds.”
The rule does not apply to work done before a chairholder came to Canada. However, some of the questions about data from Dr. Pruitt’s lab relate to his current work.
Dr. Laskowski said in an interview that she realized something was amiss when a group of researchers told her they had detected unusual patterns in the raw data Dr. Pruitt provided for the 2016 study, which was published in The American Naturalist.
Those data were obtained by timing the re-emergence of spiders that had been deliberately frightened into hiding during a behaviour experiment. Dr. Laskowski said that when she went back to the data, she found many of the measured times were identical numbers repeated again and again. When she followed up with Dr. Pruitt, she said, he told her the repeats were due to spiders being timed in groups. However, she said, further analysis showed that didn’t account for the patterns she saw. This included identical sets of numbers in different parts of the data that had no logical reason to be related. Elsewhere, some numbers appeared similar except for a difference of a single digit.
She said she lost confidence in the data and recommended the paper be retracted. Dr. Pruitt agreed, she said. She then found similar problems with the data underpinning two other papers she published with Dr. Pruitt. One of those has also been retracted, and Dr. Laskowski and other collaborators have requested the other be retracted also.
Dr. Laskowski added that before the problems turned up, she had no reason to suspect anything was wrong based on her trust in Dr. Pruitt’s track record as a productive and innovative researcher.
“He’s prolific. He’s a fast, hard worker who gets things done quickly,” she said.
In a statement posted online by Ambika Kamath, a postdoctoral researcher at the University of California Berkeley, five collaborators and former students jointly said that, over the past two weeks, they have been “grappling with this professionally and emotionally fraught situation." The group, which includes Dr. Kamath, added that they had painstakingly been reviewing "every piece of data we’ve ever been handed by Jonathan Pruitt. We have been working day and night to uncover all irregularities and communicate with the relevant journal editors.”
Another co-author, Daniel Bolnick, who is an evolutionary biologist at the University of Connecticut and editor-in-chief of The American Naturalist, is maintaining the spreadsheet of publications that have been retracted or are suspect because of data from Dr. Pruitt.
He said that he has been in contact with Dr. Pruitt, but has not yet heard a satisfactory explanation for the problems in his data. He added that while the issues are significant, they would not necessarily be obvious to a collaborator or peer reviewer who wasn’t specifically looking for them.
“I place no blame on the co-authors for missing this," he said. “Probably every animal behaviour department in North America and around the world is having conversations this week about how we can do better.”
Our Morning Update and Evening Update newsletters are written by Globe editors, giving you a concise summary of the day’s most important headlines. Sign up today.