Do you know what research misconduct is, and would you report it if you suspected it?
These deceptively simple questions reflect two pillars of research integrity: recognizing fabrication, falsification and plagiarism (and other untoward practices) and holding those who commit such acts accountable.
Yet more than 200 recipients of National Science Foundation (NSF) graduate research fellowships gave such troubling answers to these and other questions via an anonymous online survey that the authors of a paper deemed them “astonishingly uninformed” and in need of better training that uses real-world examples of unallowable behaviors.
Previous research has “contended that the ultimate responsibility to uncover misconduct rests on individual scientists, and indeed, whistleblowers have been the most common way prominent research fraud cases came to light,” the authors wrote. Yet only 30.7% of 244 fellows surveyed in 2019 said they would report a researcher suspected of misconduct; 60.7% (148) didn’t know if they would.
Additionally, when asked if they had heard of research misconduct in their field in the past five years, across the board, 63% of fellows said they had not, triggering the authors’ “astonishingly uninformed” comment.
The open-access paper, “NSF Fellows’ perceptions about incentives, research misconduct, and scientific integrity in STEM [science, technology, engineering and math] academia,” was published April 7 in Nature Scientific Reports.[1] It was authored by Siddhartha Roy and Marc A. Edwards. Roy, formerly with the Department of Civil and Environmental Engineering at Virginia Tech, where Edwards is a distinguished professor, is an environmental engineer and research associate at the University of North Carolina-Chapel Hill Water Institute.
Their paper “provides the first ever snapshot of perceptions about academic cheating and research misconduct amongst [a] high-performing group of researchers,” wrote Roy and Edwards, adding “relatively little data” exists about this group and their thoughts on these subjects.
Overall, the responses “cast doubt on the quality and effectiveness of scientific integrity trainings being offered nationwide to engineering graduate students in promoting ethical awareness and behavior,” they wrote.
The survey asked about “cheating, research misconduct, formal integrity training and ethical environments, as well as the overall positives and negatives of academia. NSF’s definition of research misconduct, i.e., the ‘willful fabrication, falsification, plagiarism, and other questionable practices,’ was displayed before survey respondents answered questions on the topic,” the authors explained.
Responses came from 133 and 111 fellows who had studied civil and environmental engineering or computer science and engineering, respectively—“two broad STEM disciplines that have transformed society but currently face concerns about ethics and high competition for faculty positions,” according to the authors. Interestingly, six (2.5%) fellows “confessed to lying in their survey responses,” and 98 or 40.2% said they were “tempted to lie.”
The group included older and more recent investigators; fellowships were from 2002 to 2007 or 2012 to 2017. Respondents completed the survey from February to May 2019. Additionally, 56.1% (137) were in graduate school at the time, while 43.9% (107) had graduated. A little more than 50% were “employed in academia as graduate students, postdocs or untenured faculty (50.8%) and tenured/tenure-track professors (20.1%).”
Tenure Pressure, ‘Laziness’ Among Factors
As noted earlier, when asked if they had knowledge of misconduct cases in their field in the past five years, just 89 (36.5%) said yes, and the authors found a significant association between knowledge and “academic stage”—“half of tenured/tenure-track professors … reported knowledge compared to less than one-third of graduate students or non-tenure track professionals.”
In addition, nine (4%) “confessed to participating in research misconduct,” and 29 (11.9%) “had first-hand knowledge of misconduct by colleagues in their research group, department or field.”
When asked to select “factors that contribute to misconduct or fraud” (more than one choice was allowed), fellows listed the following:
-
“Promotion and tenure pressures” 89%
-
“Funding hyper-competition” 67%
-
“Desire for fame” 56%
-
“Belief in one’s theory” 45%
-
“Laziness” 41%
Other responses indicated that 88.9% of fellows (217) said they “would not engage in misconduct (i.e., fabricate or falsify data) to gain funding, win scholarships or publish in high-impact journals,” another 10.7%, or 26, were “unsure.” The survey also looked at the impact of fellows’ superiors. “If pressured to engage in research misconduct by an advisor,” 7.4% (18) said they would, 37.5% (87) were unsure and 56.9% (139) would not.
The authors cautioned that NSF fellows “are also a group least subject to financial pressures during graduate school due to NSF funding and, therefore, perhaps more likely to accurately describe dominant incentives and external pressures, which may be worse for the typical graduate student.”
Turning to reporting and accountability, the authors found, as noted, only 30.7% “would report another researcher if they suspected misconduct.” Nearly 60% were unsure, and 8.6% said they would not. However, only 4% of those identifying as women said they would not report it.
This possible inaction, the authors wrote, “is probably not surprising given that academics usually have no incentive beyond curiosity, self-interest or a sense of duty to investigate research misconduct.”
In addition, “the repercussions of exposing unethical behavior are potentially catastrophic for whistleblowers, as journal articles, grant applications and awards are anonymously reviewed by peers and severe mental health problems can result from academic shunning and retaliation,” the authors said.
Roy and Edwards also noted that “the incentives for departments and universities where unethical professors bring in large amounts of funding can create conflicts of interest and should be considered.”
Often Training Was Online Only
Is misconduct inevitable? Fellows were split on this question.
Asked to estimate the percentage of researchers who would “succumb to pressure and commit misconduct at least once in their careers,” 5.3% estimated 75% to 100% would, while 61% estimated more than 10% would.
Survey respondents were also asked to identify appropriate punishments for those who commit research misconduct. While they endorsed “public retractions and corrections of the scientific record, firing or revoking of faculty tenure” and “a permanent public record of the misconduct,” only half said research in which the public was harmed should result in “charges or a criminal investigation.” Disagreement about “what constituted distortion of the scientific record” was also evident in the survey answers.
Other questions addressed integrity training; fellows indicated they had attended online sessions, university courses and workshops. Of note, 62.7% participated in online training offered by the Collaborative Institutional Training Institute (CITI). The authors said more than 2,200 institutions offer only CITI training.
The authors called it “concerning" that 54.1% of the surveyed fellows thought their training had no effect on their “ability to handle ethical dilemmas.” Forty-four percent felt more prepared.
Regarding their own behaviors outside of strictly defined research misconduct, 39 (16%) said they had “cheated in college and/or graduate school,” while 205 (84%) said they had not.
Seventy-six (31.1%) reported having “seen their graduate peers cheat.” Copying assignments (81.6%), plagiarism (47.4%) and “using online solutions” (36.8%) were the three most common types of cheating.
“The top two reasons Fellows offered for committing academic cheating or considered a motivation for their peers cheating were good grades (e.g., ‘afraid of bad grades—ashamed of having done so!’) and having less time (e.g., ‘felt too busy, had to cut corners to get everything done’). In one department, it was asserted that cheating was the norm (i.e., ‘it [was] unusual if you DON’T have the homework solutions ahead of time’). In another, ‘getting at least the A or B grade [was] required to continue in the program,’” according to the authors.
Motivations to Cheat Must Be Understood
Other reasons cited were “the drive to stay competitive (e.g., ‘I felt that it was a gray area and that I wanted to have a leg up on my classmates’), the advanced nature of graduate-level classes, and preference to do research over classwork (e.g., ‘classes are a waste of time, would rather do research’), were less prominent but still notable factors … motivating Fellows and their graduate peers to cheat. Altruism (e.g., ‘I was enjoying working with friends and wanted to help them’) was also mentioned.”
More immediate efforts to “reduce cheating and research misconduct ... should consider both individual motivations and academic pressures. Pressure to get promotion/tenure was top-ranked by Fellows as possible motivation behind unethical behavior, which is consistent with recent findings on researcher career stage being a predictive factor for journal retractions that mostly result from scientific misconduct. Integrity training should likely include real world and field-specific case studies and instruction rooted in human nature and organizational psychology,” the authors wrote.
Additionally, they recommended that “educational psychologists, moral psychologists, neuroscientists, behavioral economists and legal scholars” work together to “(a) design ethics training and interventions that reduce occurrence of academic dishonesty and research misconduct, (b) isolate institutional and field-specific factors that impact motivation and likelihood of misconduct, (c) study the relationship between individual personality traits vis-a-vis academic cheating and misconduct, and (d) formalize and refine conflicts of interest, penalties and reparation processes for misconduct.”
In the longer term, the authors suggested that surveys like theirs should be repeated every 10 years.