Irreproducibility has long been a problem in scientific inquiry. While this can sometimes be blamed on a lack of skill or poor methodology, it can also be due to cheating and forgery that should be addressed aggressively. In a recent article in Nature, titled “Stop Ignoring Research Misconduct”, authors Donald S. Kornfeld and Sandra L. Titus describe the reasons for irreproducibility and ways the scientific community can combat the practices that lead to it.
According to the authors, “faulty research practices and fraud” are the two reasons for irreproducibility. In all of our optimism, the scientific community believes that irreproducibility is not due to misconduct, like fraud and fabrication. This has been declared by leaders in science, including the US National Institutes of Health. Kornfeld and Titus argue that this willful disregard of deliberate misconduct is not only a squandered opportunity for education and outreach, as it also creates an environment in which such behavior is not only condoned, but rewarded.
The authors cite a study of more than 1,000 post-docs, in which more than a quarter of respondents state that they “would select or omit data to improve their chances of receiving grant funding”. While the authors admit that that there are many cases of irreproducible research that do not involve misconduct, fraud should be taken seriously and addressed directly in the research community.
The authors recommend what they describe as five key approaches to reduce misconduct:
Authorities should acknowledge that deliberate misconduct is an important contributor to irreproducibility.
Mentors should be evaluated to assure quality; those who contribute to misconduct should be penalized.
Institutions and government agencies should have procedures to protect whistle-blowers from retaliation.
Senior faculty members who are found guilty of misconduct should face severe penalties.
Institutions that fail to establish and follow policies and processes to prevent misconduct should be sanctioned.
While all NIH funded trainees must undergo responsible conduct of research training, defined as “fabrication, falsification, or plagiarism,” Kornfeld and Titus argue that trainings should instead address “psychological factors” that lead to misconduct, which differ depending on the individual’s role in the research.
Below is a summary of the common pitfalls identified by Kornfeld and Titus and their recommendations for each type of investigator:
- Trainees: Trainees resort to misconduct as a result of fear of disappointing their mentor or due to a lack of quality mentorship. The authors suggest that mentors should carefully balance their mentee workload with their other responsibilities and limit their number of mentees to ensure that each mentee is receiving adequate attention. They also suggest that trainees complete anonymous questionnaires about their mentors to appropriate college authorities and funding agencies. Ultimately, senior staff must take responsibility for their trainees.
- Support staff: Support staff may resort to such practices to increase their income or decrease their workload. The authors recommend treating the support staff with the respect they deserve as vital to the project goals, informing them of the study’s purpose, and explaining the dangers of deliberate misconduct.
- Senior researchers: Senior researchers often perceive a low risk of exposure and subsequent punishment, as the scientific community does not consider deliberate misconduct to be a top priority in reducing irreproducibility, and few cases are referred to the US Office of Research Integrity. Because retaliation is currently an unfortunate reality for many whistle-blowers, Kornfeld and Titus recommend having clear policies regarding allegations, as a well as a respected faculty member or administrator serve as a research integrity officer to protect those who come forward.
- Institutions: According to Kornfeld and Titus, institutions should “build a culture and infrastructure that encourages integrity.” Random audits, date-stamping data systems, and well-enforced policies are just some of the author’s suggestions of ways an institution can mandate compliance; however, a culture and open support of such policies and security features will help tremendously in preventing misconduct. The authors believe the institution must take responsibility for its researchers and ensure that adequate policies, procedures, and educational efforts are in place to prevent misconduct.
Ultimately, Kornfeld and Titus hope that the scientific community will confront the issue of deliberate misconduct, which accounts for more instances of irreproducibility than we would like to believe.
The title quote, “Hoaxing, Forging, Trimming, and Cooking” was taken from Charles Babbage’s 1830 work, Reflections on the Decline of Science in England.
 Donald S. Kornfeld and Sandra L. Titus. “Stop Ignoring Misconduct,” Nature 537, 29-30 (2016).
 Kornfeld & Titus, “Stop Ignoring Misconduct,” 29.
 Francis S. Collins and Lawrence A Tabak. “Policy: NIH Plans to Enhance Reproducibility,” Nature 505 (2014), 612–613.
 Susan Eastwood, Pamela Derish, Evangeline Leash, and Stephen Ordway, Stephen. Sci. Eng. Ethics 2 (1996), 89–114.
 Kornfeld and Titus, “Stop Ignoring Misconduct,” 30.
 Collins and Tabak, “Policy: NIH,” 612–613.