By: Paul S. Thaler, Paul E. Simon, and Sydney Smith Forquer
In recent weeks, members of the press and bloggers have paid significant attention to the resignation of former Stanford University President Marc Tessier-Lavigne. The articles feverishly report on the investigation into allegations that Dr. Tessier-Lavigne engaged in research misconduct. Many of the articles covering this so-called “scandal” imply that he is somehow responsible for bad or fraudulent science—despite the fact that the scientific panel tasked with reviewing the 12 publications at issue concluded that Dr. Tessier-Lavigne did not engage in research misconduct and did not have actual intent or knowledge or act with recklessness relating to any research misconduct in which his lab members may have been engaged. This public reaction underscores many of the more troubling—and potentially dangerous—reactions to even the mere allegation of research misconduct.
What is missing throughout the case (and the associated press coverage) is any explanation of why the university ignored the established process for handling allegations of research misconduct. The university instead conducted an investigation using a “special committee,” which, although purported to have followed the standards promulgated in the federal research misconduct regulations and the corresponding Stanford policy, expressly disclaimed making a determination pursuant to those regulations and policy. Well-established procedures introduced by Congress more than thirty years ago and mirrored by all Public Health Service-funded institutions, including Stanford, provide an appropriate definition of standards for research misconduct investigations while protecting the interests of the respondent, complainant(s), and science itself.
In fact, we are in the dark as to what process the university used. There are set rules and protections in the research misconduct process, and if the special committee followed these rules, then why was the public made aware of the existence of the process at its outset? It would be a clear violation of the confidentiality provisions that exist to protect both those who make research misconduct allegations and those who are defending themselves against the allegations. If this was outside of the research misconduct process, it is reasonable to ask why when such a process exists expressly for this type of investigation. And if there was a parallel research misconduct process that remained appropriately confidential, why was Dr. Tessier-Lavigne required to go through a second investigation for the same allegations? That smacks of functional res judicata violations.
Of course, research misconduct—in the form of fabrication (making up data or results), falsification (changing or omitting data or results), or plagiarism (appropriating another scientist’s ideas or words without credit)—must absolutely be taken seriously. Indeed, the Public Health Service regulations applicable to most federally-funded research were introduced to protect science following the discovery of several high-profile, disturbing instances of scientific fraud. The express purpose of the federal regulations is to protect public health and safety, promote research integrity, and conserve federal funds. This purpose is mirrored in the corresponding institutional policies that entities receiving federal funding must have in place as part of their assurance to the government.
However, the purpose of these policies and regulations is not to publicly shame or punish even those scientists who are ultimately found to have engaged in research misconduct. For this reason, research misconduct proceedings are intended to be protected by confidentiality, with the disclosure of the identity of respondents limited to those who “need to know” in order to conduct a thorough and fair investigation. Any transparency into the identity of respondents in research misconduct processes is meant to occur when (and only if) the federal Office of Research Integrity (ORI) makes a federal finding of research misconduct, which is then published in the Federal Register. The confidentiality requirement is meant to encompass even the investigation reports themselves; thus, the disclosure of the scientific panel’s report concerning Dr. Tessier-Lavigne is troubling. The very fact that public shaming, such as what has surrounded Dr. Tessier-Lavigne, continues to happen when information becomes public contributes to an ongoing stigma. This stigma understandably makes scientists reluctant to correct or retract publications even when they become aware of errors that might be entirely benign or are due to honest mistakes. It was especially alarming to see press coverage from the very beginning of, and perhaps even before the initiation of, Stanford’s investigation. Confidentiality in the research misconduct process is partly to allow for the possibility that the respondent is not found to have committed research misconduct. And, of course, that is actually what happened in this case!
While it remains unclear what process the special committee and its scientific panel followed in reviewing this matter, the scientific panel stated that it followed the federal research misconduct regulations and the associated Stanford policy in reviewing the research misconduct allegations. Notably, pursuant to both the federal regulations and the Stanford policy, the burden on an institution for finding a scientist engaged in research misconduct is remarkably low, a mere preponderance of the evidence, meaning that an investigation must only determine that it is “more likely than not” that the scientist is responsible for research misconduct. This is the same standard of proof for a civil trial, although what is at stake for the scientists accused is far more significant than just a sum of money—often, their very ability to continue in their chosen career is on the line. This is despite the fact that even a civil trial provides more safeguards, with a civil defendant provided with discovery, a transparent window into everything that occurs throughout the trial, and an ability to cross-examine those providing evidence against them. In contrast, a scientist accused of research misconduct is only given access to the record at the end of the investigation and has no opportunity to confront those providing evidence against them. Even though this remarkably low burden could not be met as to Dr. Tessier-Lavigne, his treatment in the press is unsettling.
Indeed, the scientific panel’s report does not merely state that a preponderance of the evidence weighed against a finding of research misconduct but went much further than that: “At no point in its work did the Panel encounter evidence to conclude that Dr. Tessier-Lavigne acted inappropriately to manipulate research data within these or any other papers, nor did it encounter evidence to conclude that he knowingly countenanced others doing so” (emphasis added). The scientific panel expressly disclaimed meeting its burden as to finding even reckless intent by Dr. Tessier-Lavigne, let alone meeting the burden of proof for all required elements of a research misconduct finding.
Critically, this extremely low burden of proof is never met in most research misconduct matters. As reflected in ORI’s most recent three annual reports to Congress (covering October 1, 2019 through September 30, 2022), ORI received 703 research misconduct allegations, which resulted in a determination that only 128 cases should be opened (only approximately 18% of allegations rose to a level sufficient to open a research misconduct matter). Over the same period, ORI closed 245 cases, out of which only 26 resulted in findings of research misconduct (only approximately 10% of research misconduct cases resulted in a finding that the scientist engaged in research misconduct). These numbers underscore a fundamental truth: scientists, like the rest of us, are human and, as humans, are fallible and subject to making mistakes.
As in the McCarthy era when many lives were ruined based on mere allegations that individuals were Communist, research misconduct allegations are sometimes made not based on a good faith belief that the science was wrong, but for purely political or petty reasons. This, too, can be seen in the number of allegations that never result in the research misconduct process beginning or that conclude during the assessment or inquiry stage of an institutional proceeding instead of making it to an investigation.
Concerning Dr. Tessier-Lavigne, although 12 papers were investigated, the scientific panel concluded that he was not in a position to detect the alleged errors in seven of them, for which he was not a principal author. Even with respect to the other five papers, the scientific panel concluded that Dr. Tessier-Lavigne never acted inappropriately and did not countenance lab members engaging in research misconduct. Significantly, the scientific panel’s report expressly states that “[b]ased on the facts reviewed, it would not be reasonable to expect Dr. Tessier-Lavigne to have identified these instances of research data manipulation prior to or at the time of the respective papers’ publications . . . . Indeed, several of these manipulations went undetected by the scientific public even with the use of cutting-edge image analysis tools available fifteen to twenty years after the papers’ publications” (emphasis added).
In short, the scientific panel found no fault with Dr. Tessier-Lavigne as a scientist. At most, it found that he did not take sufficient steps to correct the record after becoming aware of concerns. Nevertheless, even in these instances, the fault lies not on Dr. Tessier-Lavigne—who acted appropriately by contacting journals—but on the journals themselves, who failed to take the action the scientific panel apparently preferred. The scientific panel suggested that Dr. Tessier-Lavigne should have continued to press the journals; however, as the publishers of the papers, it is the journals who decide whether or not to take any action. Neither Dr. Tessier-Lavigne nor any other scientist who requests a correction or retraction has control over the journal and should not be berated for an alleged “failure” to force the journal to take action.
In our experience, having defended hundreds of scientists accused of research misconduct, many scientists who face such allegations never did anything wrong or made minimal errors. Nobody is ever entirely mistake-free. A scientist who makes a mistake, even as the result of sloppiness, should not have their reputation torn to pieces and their career harmed. Until the stigma of bad science is separated from the very notion of a correction or retraction, scientists understandably remain reluctant to seek such measures. And the level of potential damage to their careers due to public awareness of even an allegation of research misconduct is reason enough that a much higher burden of proof should apply to research misconduct matters.
The panel noted that Dr. Tessier-Lavigne was “open, cooperative, and professional throughout these proceedings.” That is not the behavior of a “bad” scientist or one with something to hide. A preeminent, distinguished scientist such as Dr. Tessier-Lavigne should not be shrouded in “scandal” when exonerated of research misconduct allegations simply because in hindsight, others think he should have done more to convince the journals to take action. When otherwise good scientists are unduly and unreasonably punished by the research misconduct process and subjected to public ridicule, it is not only their careers that are harmed, but science itself—an outcome that helps nobody and is itself at odds with the purpose of the research misconduct regulations.