The Dilemma of Reporting Your Ph.D. Advisor for Falsifying Data 

Not long after starting as a biomedical engineering Ph.D. student at Vanderbilt University, June (names have been changed for this story) started losing trust in her boss, the principal investigator (PI) of the lab. In one instance, he had encouraged experimental surgeries on mice without proper ethics approval. Even though a trained medical doctor would perform the surgeries, it still struck her as reckless. What bothered her more, though, was that she couldn’t replicate published data related to the project she inherited when she joined the lab. 

“It was bad science,” she said. The data wasn’t falsified, but it was sloppy enough to require a correction, so she brought it up with the PI. Soon, she received an email from him, with all members of the lab cc’d, outlining an ultimatum: she could either spend the next three months looking for a new job or try to solve the problems she uncovered. At the end of the three months, the lab members would vote on whether she can stay. 

The threat spooked her, but as a high-achieving engineering student, she was unaccustomed to giving in. Night after night, she worked in the lab. She consulted with an experienced chemist when she ran into roadblocks. In the end, when she presented a solution to the PI there was no final vote or mention of the ultimatum, and the PI seemed happy to publish her work. However, she was left believing he would skirt ethics to achieve results, and he didn’t respond well to scrutiny. These beliefs would guide June’s actions when she made a more damaging discovery later in her Ph.D. In the turmoil of current events, we hear that silence is complicity, but how does this idea apply to an abstract concept like protecting the integrity of science? What if reporting misconduct is likely to lead to retaliation by the wrongdoer and disregard by the systems that should help? 

From the start, June and her PI had had a tumultuous relationship. They argued often, but he also told her he wanted her to succeed because they were both from South Korea. She knew his family and babysat his kids. In late 2013, the lab was low enough on money that he borrowed from other professors. June sensed things were getting desperate, so she helped prepare a grant proposal for the National Science Foundation (NSF). In the proposal, the PI included data that seemed odd. Based on the type of experiment, only June and one other person could have produced the data. Did the other student do an experiment she wasn’t aware of? She asked around. According to everyone else in the lab, that experiment had never happened. 

Speaking to a group of friends about what she found, somebody suggested that falsifying data is not something that people do only once. “I’ll go through all the copies of grants that the PI had shared with me over the past few years and see if there are other similar incidents,” June thought. What she found dismayed her. Two additional cases emerged in which he used a single figure with the labels changed to represent different experimental results. In another, he completely fabricated data from a mouse surgery. 

The guidelines of the Office of Research Integrity define research misconduct as plagiarism, falsification of results, or fabrication of data. Fabrication of data involves making up data, whereas falsification is the misrepresentation of data, such as manipulating or omitting results. June sensed what she discovered fell under the definition of research misconduct. 

June turned to former students for advice. One student June spoke with had also found evidence that their PI had falsified data. The student “was angrily shot down,” when she mentioned it to the PI. June struggled with whether to report her PI to the university. On the one hand, she knew that falsification went against the integrity that she believed existed in science research. On the other hand, she knew how volatile her PI could be when confronted. Would reporting him to the university derail the three years she invested in her Ph.D.? 

She felt the safest decision was to confront her PI one-on-one after she graduated. “I felt like I was a hostage until he let me graduate,” June said. She made recorded the evidence of falsification, set them aside and moved on with her work. 

June’s intuition that there would be negative consequences for reporting her PI was accurate. In one publicized case of misconduct at the University of Wisconsin, several students who reported their boss didn’t finish their graduate studies and left the university. The role of power is paramount to these cases: a study reported that younger researchers are less likely to report misconduct due to their limited power and social capital. Power dynamics may also result in fewer women reporting misconduct. Only 17 percent of Ph.D. student whistleblowers were satisfied with the outcome of reporting misconduct. 

The number of reports of misconduct made to the NSF is rising. In a survey of scientists, only around two percent of scientists admitted to faking data. However, one-third of scientists agree they’ve committed “questionable research practices.” Approximately 14 percent of scientists surveyed say they’ve seen colleagues committing misconduct. These data suggest that many cases of research misconduct go unreported. More bluntly, Dr. Peter Charles Hoffer, a researcher of academic fraud, wrote, “it’s like cockroaches. For every one you see on the kitchen floor, there are a hundred behind the stove.” 

There are obvious incentives to fake data. Publications are like a currency that scientists use to compete for grants and promotions–the more you have, the further you'll go. To some, cheating is a steam outlet in a rapidly boiling pressure cooker. It is twisted, and detracts from the honest successes of their peers, but is not surprising. 

In autumn of 2015, June was mentoring Ethan, a new graduate student. He noticed immediately that the PI was unpleasant to be around and started asking June questions. What is he like with others in the lab? Is his temper a problem? At this point, her silence about the falsification had taken a toll on her health. “It definitely was eating away at my soul,” she said, and she worried about Ethan and what might lie ahead for him. June connected him with former students who might feel comfortable speaking freely. Over email, one former student mentioned that the PI “might ask people to do ethically questionable things.” The vagueness of this response only piqued Ethan’s questioning. 

Eventually, June told him what she knew, and Ethan decided to leave for another lab. He didn’t slip out silently, however. The Director of Graduate Studies noticed that students in this lab were unhappy, and he pressed Ethan for reasons. Ethan spilled the beans about everything. He told the Director about June's evidence of research misconduct. When telling this story, June laughs at the absurdity of it, but it would come to change the course of her graduate experience. Before long, a group of Deans from Vanderbilt contacted her. 

“At first, when the deans came to me, I was like holy shit, this is not how I intended this,” June said. “But, I confirmed that I was willing to cooperate even though I was scared.” 

Over the next month, however, June heard nothing from the administration. Their disregard troubled her. Reaching back out to the deans, she asked for an update and also requested legal counsel. The school’s lawyers were out of the question, she was told, because of conflict-of-interest. All law firms she contacted in Nashville claimed an affiliation with Vanderbilt University. Nobody would take her on as a client. It was several more weeks before a senior dean in the engineering department contacted her to discuss the allegations. 

When research misconduct is reported, the university carries out an initial inquiry to determine if the evidence warrants a full investigation. If so, the university must also report the incident to the federal agencies that funded the grants in question. These agencies conduct a separate investigation and may take disciplinary action. 

If you spot the conflict of interest here, you’re not alone. 

When a university performs the initial inquiry, they benefit (avoiding the negative press and retaining funding) from there being no misconduct. William J. Adams, a Professor of Mathematics at Pace University, wrote: “academe often discourages victims from seeking justice, and when they do, tends to ignore their complaints—a kind of scholarly ‘don’t ask, don’t tell’ policy.” 

When June met with the senior dean of engineering to discuss the allegations against her PI, he told her she didn’t have a legal obligation to report anything officially unless someone was harmed physically. “I was disturbed and shocked and really scared because how do I trust a system like that?” June said. When the PI learned that Ethan was moving labs, he sent an email notifying the research group of the change. But that wasn’t all. He also revealed that Ethan had cystic fibrosis, which was true but confidential. June was offended and forwarded the email to Ethan and alerted the director of graduate studies. 

The next morning, she had a regular one-on-one meeting with the PI where they typically discussed her research. At the end of an otherwise ordinary meeting, she told him she thought it was wrong to have shared Ethan’s medical information. He stared at his computer monitor and didn’t respond. June went on to request time off from work. The combined stress of the decision to file complaints of misconduct against her PI, while also appeasing him enough to graduate was growing dangerous. She told him that her health was deteriorating. What she didn’t share was that she had begun to have suicidal thoughts. 

The PI was surprisingly supportive, June remembers, and she left his office relieved. Later that day, however, she answered a call from him. He was angry. An email from the university administration had arrived in his inbox, scolding him for sharing Ethan’s medical information. June admitted she told them. 

After receiving a wrist slap for sharing confidential medical information, what does the PI do? He doubles down. He demands that June bring him her medical records, going so far as to cc the Department Chair and the Director of Graduate Studies. In his message, he claimed that June and her mental health issues were the basis of all the problems in the lab. He seeded this idea with lab members: later in November, a colleague approached June and asked if her mental health was affecting her decision to report their PI. 

Not only is retaliation against whistleblowers rampant, but June experienced a particularly devious variety: the use of mental health as a weapon. In both corporate and academic settings, the wrongdoer will often attempt to discredit a whistleblower’s mental wellness. Famously, Daniel Ellsberg, the economist who leaked the Pentagon Papers, was the victim of such an attack. Nixon operatives broke into Ellsberg’s psychiatrist’s office, hoping to weaponize his medical records against him. Instances of depression and suicidal thoughts, much like June experienced, are common because of the stress brought on by whistleblowing. 

Of course, the message that June bubbled up to the dean–that he mishandled health records–wasn’t her biggest bombshell. The PI’s attempt to undermine her by asking for her health records was the impetus to file the research misconduct allegations she had already prepared. 

Even though June saw the falsification as a betrayal of the integrity of science, “What really pushed me was this medical information mishandling. It was the last drop that I could take,” she said. She admits that if the PI had not attacked her personally, she probably would have handled things behind closed doors. A study of medical doctors found that patients were more likely to submit malpractice lawsuits against doctors with poor bedside manner. The PI continuously showed poor bedside manner. The likelihood of an individual to whistle-blow is associated with the seriousness of wrongdoing, and whistleblowers perceive wrongdoing as more serious if it occurred frequently. In June’s case, the mishandling of her medical situation violated her personal code of ethics. 

June prepared the documents and submitted them to the university in mid-November 2015. Days before Thanksgiving, at 11 am on a Tuesday, Vanderbilt University Police entered the PI’s office and seized his computer. Once again, the PI suspected June. This time, however, he apologized and asked her to meet in person. 

In his office, he confessed that he had been investigated for misconduct before, having plagiarized text in three NSF grant proposals. June outlined the evidence that she submitted to the University for the inquiry. His apologies struck a chord with her; she began to worry about his family. After everything was in the open and June was preparing to leave, he begged for her help. 

“Can you please help me get out of the situation without regard to your ethical standards, and I will help you graduate in whatever way you like,” he told her. She only realized what he was trying to do after she left his office. It was a bribe. She severed all communication with the PI in early December 2016. 

June kept meticulous notes throughout the process. To prepare for questions the inquiry committee might ask, she was reviewing her coworkers’ lab notebooks one evening when she came across one last transgression. One of her claims suggested that the PI fabricated a mouse surgery, and she noticed that someone had written over a similar experiment to make the record look like the surgery had taken place. The writing was in different color ink and was strikingly similar to her PI’s. June forwarded this finding to the inquiry committee and waited to hear back. 

Over a year later, the National Science Foundation’s Semiannual Report to Congress contained an anonymized summary of a research misconduct case. The summary was titled “Assistant Professor who previously committed research misconduct falsified data.” It outlined falsified data in two figures submitted in grant proposals to the agency and mentioned the falsification of data records in a lab notebook, clearly referring to June’s PI. A link to a past misconduct case, which occurred at a “Tennessee university”, was included. The summary recommended that the “NSF make a finding of research misconduct and debar the PI for five years.” It also recommended that the NSF certify the integrity of his proposals for ten years. The summary says that the PI resigned from Vanderbilt, but the university website suggests he kept an academic position there. He now holds a professorship in South Korea. 

As for June, she defended her Ph.D. (the university banned the PI from attending her defense) and graduated. She moved to Boston and now works at a biotechnology company. Although she is still skeptical of men who are more senior than her in the company, her distance from Nashville has helped her move on. When considering others who might be in her position, she hesitates: “I don’t know what the ethically realistic thing to do is,” June said. “You either have to choose to be realistic and keep your mouth shut and have your career not tainted or choose to be stupid and idealistic and file it officially.”