Report on Research Compliance 17, no. 8 (August 2020)
Despite data falsifications that were severe enough to require four retractions or corrections and which led to his demotion at a university and a redo of experiments, a postdoctoral researcher escaped any sanctions by the National Science Foundation primarily because, in NSF’s view, its Office of Inspector General (OIG) took too long to close the case. However, there is no statute of limitations that applies in research misconduct cases, RRC has learned.
NSF has the authority to impose a range of sanctions, from lifetime debarment to training and integrity certifications, against researchers who commit misconduct, defined as fabrication, falsification and plagiarism. With some variation, these are the same that the HHS Office of Research Integrity (ORI) has at its disposal when it makes a finding involving research supported by Public Health Service agencies, such as NIH.
The unidentified researcher (his university is also not identified) is just one of two cases of research misconduct that NSF considered for action from Oct. 1, 2019, to March 31 of this year, according to OIG’s recent semiannual report (SAR) to Congress.
Historically, NSF OIG does not identify the names of sanctioned investigators nor their institution—even when debarment is imposed. Debarment and exclusion are black marks that generally mean no federally supported institution can do business with them (at least for the period of the sanction), something organizations need to know. This information can be obtained through authorized access to a nonpublic database, but the general research community and the public (and news media) won’t be able to learn about such principal investigators (PIs) or their institutions.
The names of NSF-funded researchers who are placed under supervision or have other less severe actions taken against them, such as being banned from peer review and advisory panels, are generally never known outside of the institution; there is no such database (public or otherwise) where their names would be recorded.
In response to questions from RRC, OIG said it does not identify researchers because of federal laws.
“The identities of these individuals, and information that potentially could reveal their identities, is protected from public disclosure by laws and regulations pertaining to personal privacy, such as the Privacy Act and the Freedom of Information Act,” OIG said.
However, the same laws apply to ORI, and it informs researchers that it will be publishing all the details of misconduct findings in the Federal Register.
OIG first described the case of the postdoctoral researcher in its previous SAR to Congress that covered the period from April 1 to Sept. 30, 2019. “A postdoctoral researcher falsified data in multiple papers and progress reports to NSF. The university investigation committee determined that four articles containing the researcher’s data required either correction or retraction. The committee determined the researcher had a ‘practice of searching his notebook and scans for bands with the same intensity patterns that he could use as surrogate’ figures as they were more aesthetically pleasing. The committee concluded the researcher committed research misconduct by intentionally fabricating data. They recommended the researcher be demoted, write a report outlining specific guidelines for data presentation from journals specific to his discipline, attend ethics and RCR [responsible conduct of research] courses, repeat experiments for the figures in question, and be subject to a strict mentoring program,” the report said.
OIG “concurred with the university committee that the researcher committed research misconduct and recommended NSF require him for 2 years to submit certifications and assurances, bar him from serving as a reviewer, advisor, or consultant for NSF and provide proof that he completed the university mandated training.”
Certifications and assurances are attestations that the researcher and/or the institution would make to NSF that “all proposals or reports” and other documents submitted do not “not contain plagiarism, falsification, or fabrication.”
In the newest report, OIG briefly related NSF’s decision. “In the case of a post-doctoral researcher who fabricated data in multiple publications sponsored by NSF, the agency chose to take no action due to the extended period of time between the dates the researcher committed the acts and the date NSF received our report.”
In contrast, ORI has frequently closed misconduct cases with findings and sanctions six years or more after the misconduct occurred. For example, in November 2019, ORI made a misconduct finding and imposed a four-year debarment against a researcher who falsified data in a 2008 paper and her 2010 thesis.
RRC contacted both OIG and NSF for more information, such as how much time had elapsed. OIG called the timeframe “atypical” but indicated it was not required to complete its investigation by any particular time, and explained why the case took a long time to close.
“OIG opened the case in May 2013, and NSF received the report in June 2019,” OIG told RRC. “While there is no statute of limitations for administrative research misconduct cases, the timeframe in this matter is atypical. Contributing factors involved staff turnover and shortages, case complexity, and competing priorities. We are unable to address NSF’s decision-making beyond what we reported in the SAR.”
In its comments to RRC, NSF said that, “Given the six-year delay, there was no further reason for NSF to take additional actions. Such actions would have duplicated actions already taken or imposed by the university.”
However, as noted, NSF did not actually make a misconduct finding in this case, despite the university’s determination as well as OIG’s. NSF could have imposed additional requirements that were not imposed by the institution, that were not duplicative or were specific to NSF, such as prohibiting the researcher from advising NSF.
NSF also did not follow all of OIG’s recommendations in the second case in which it took action during the reporting period. OIG recommended that NSF debar a PI who “plagiarized text into an NSF proposal from two awarded NSF proposals that he received in confidence from a colleague.”
But it did accept other recommendations, which are to “certify compliance with university-imposed requirements and provide certifications and assurances for 4 years. NSF also barred the PI from participating as an NSF peer reviewer, advisor, or consultant for 4 years,” according to OIG.
The SAR also discusses seven completed investigations that have been submitted to NSF for adjudication. Two of the seven cases involved data fabrication; the other five involved plagiarism. In three of the seven, OIG recommended debarment from one to three years—the most severe sanction, with the exception of lifetime debarment. In one case the investigator was identified as female; four males were involved in other cases. For two cases, no gender was provided.
A graduate research fellow and an associate professor were among those OIG said committed misconduct. However, for the others, no titles or descriptions were provided other than that they were either a PI or co-PI. The cases generally show OIG working in tandem with the university, but in some instances, OIG was the first one on the scene, so to speak; whether it received an allegation isn’t made clear.
In just one instance, OIG disagreed with a university’s finding that there was no misconduct and made its own finding. It has not been uncommon in SARs for OIG to describe similar instances where it has felt a university didn’t do a thorough investigation or reach a reasonable conclusion. It’s also not unusual for a university or other organization to impose no sanction because the individual had already left its employment. In such cases, as this new report shows, OIG still advocates for NSF to impose a punishment.
One PI “accepted responsibility” for plagiarism; two others were said to offer a confession, although in one case it did not reflect the full extent of the misdeeds.