As consumers of the outputs of science, it is our trust in science that gives us confidence in the myriad of objects, technologies and experiences that are part of our everyday lives. In science, the ideal is that researchers look for the truth about how the world works in a transparent way that allows for reproducibility and that, ultimately, science is self-correcting. The direct relationship between science, our everyday lives and the slow iterative nature of the scientific process has been illustrated by Covid-19 vaccinations which were developed in record time only because of the decades of vaccine research that had gone before the pandemic. Anyone who has had the vaccine has demonstrated their trust in those decades of research. Of course, the pandemic has also brought to the fore many concerns about scientific integrity. With cases of research misconduct hitting the news headlines there is renewed interest in the causes and consequences of research misconduct.
Research misconduct wastes time and money, contaminates the published scientific record, hinders scientific advancement and ultimately risks harming us all as consumers of research. What is the true extent of research misconduct, and should we be concerned? Is research misconduct on the increase, or are we becoming better at uncovering it? These are difficult questions to answer for many reasons. In particular, because of the underlying research culture which drives misconduct and difficulties inherent in defining what it is.
While most definitions of research integrity share variations on the principles of honesty, accuracy, transparency, accountability, respect and dignity, the definition of research misconduct is less precise. Some definitions focus on fabrication, falsification and plagiarism, while others extend this core definition to include a range of additional behaviours. One grey area is around questionable research practices. These are a range of behaviours that result in the preferential reporting and publication of findings that are deemed interesting. This distorts the published record so that important negative data remain unpublished and hidden. The term questionable practices implies that these activities are not misconduct. After all, they are often learnt or expected from senior colleagues. In that sense, they are not as ‘bad’ as intentional misconduct. But, when considered in terms of impact on the integrity of the published record, the distortive effect of questionable practices has the same effect as overt misconduct. It has been argued that questionable practices are the most important issue as a threat to research integrity, not fraud.
Research misconduct is a complicated multilayered problem that is difficult to define because it encompasses a range of behaviours. But that is not the only challenge. The research culture that drives misconduct hinders research into the phenomenon as well.
At the UKRIO annual conference last year, a common theme was the need to address the current research culture that rewards individual achievement rather than good practices in science. The culture motivates individuals and institutions to self-promote and build reputations as infallible, excellent, and top-ranking. This fosters power imbalances, fear of blame and secretiveness which in turn prevent the open and transparent sharing of information about research misconduct.
Research integrity issues
Attempts to research the extent of research misconduct have found that a smaller proportion of researchers admit to committing certain types of misconduct themselves, than observing similar behaviour in their colleagues. Clearly, there are constraints to asking researchers about misconduct behaviour, even with the promise of anonymity. Journal article retractions have also been used to measure the extent of research misconduct, but this has limitations too. Articles can be retracted for a variety of reasons including honest error, and while COPE guidance recommends that retraction notices state the reasons and basis of the retraction, this is not always the case or easy to determine because of opaque wording used in retraction notices.
Research integrity issues for journals and publishers come to light via complaints and whistleblowers after research has been published. Manuscript screening processes including the use of tools like plagiarism detection software also help to detect misconduct before publication. Because journals and publishers investigate allegations of research misconduct, they possess data which, if shared, might help to understand the problem of research misconduct better. However, journals and publishers are trapped in the same culture as individual researchers and institutions. They too are motivated to convey an image of excellence and infallibility. Publishers and journals, protective of reputations, are not open and transparent about the extent of misconduct they find. The COPE Case Taxonomy provides some insight into the types of cases brought to COPE by journals and publishers and, at the last Peer Review Congress, a publisher presented an abstract on the types of research integrity issues it handled. This type of research on research misconduct by publishers is rare. The current research culture is a force that makes it impossible to understand the true state of research misconduct while guaranteeing to make it worse. That we can go about our daily lives consuming the products of research without mishap suggests that our trust in science is well-founded for now, but without significant intervention, will this still be the case in 20 years' time?
Jigisha Patel, Independent Research Integrity Consultant and COPE Council Member
- The evolving relationship between universities and scholarly publishers, COPE Seminar Keynote address 2021
- Ethical authorship versus fraudulent authorship, COPE Seminar session 2021
Submitting a guest editorial or opinion piece to COPE
We welcome guest editorials and opinion articles regarding research and publication ethics from COPE members. Please read the COPE guest editorial guidelines before submitting your article.