Research is fundamental to the process of pushing back the frontiers of human knowledge and understanding. Research helps cure diseases, tackle climate change, and understand the world around us. The UK has an enviable reputation for high-quality research, and researchers are among the most trusted groups of people in the eyes of the public. It is recognised that the vast majority of research undertaken in the UK is of high quality and high integrity.
Nevertheless, error, questionable practices, and outright fraud are possible in any human endeavour, and research integrity must be taken seriously and tackled head-on. The 2012 Concordat to Support Research Integrity provided a set of high-level commitments in this vein, but, six years on, while all the most research intensive-universities are complying with key recommendations of the Concordat, around a quarter of universities overall are not fulfilling the basic Concordat recommendation of producing an annual report on research integrity.
Compliance with the Concordat has technically been a prerequisite for receiving funding from UK research councils and higher education funding councils since 2013, but non-compliance has not led to any hard consequences. This reflects the fact that the Concordat has only high-level commitments and recommendations, meaning that ‘compliance’ is difficult to assess in practice. More broadly, there has been a lack of co-ordinated leadership to drive the implementation of its recommendations in universities, such as transparency in declaring the number of misconduct investigations carried out each year. The Concordat should be tightened so that compliance can be more easily assessed, with a timetabled route-map to securing 100% compliance. We welcome Universities UK’s plans to convene a meeting of the Concordat signatories to discuss the issues raised in our report and look forward to seeing further action in this area.
The current lack of consistent transparency means that it is impossible to assess the scale of the research integrity issue, leading to accusations that parts of the sector are policing themselves in a secretive way in order to maintain its reputation or, worse, a perception that investigations are not conducted properly in order to avoid embarrassment. Meanwhile, there is a risk that a future high-profile scandal could expose any weaknesses in this arrangement. Fraud appears to be rare, but the number of institutions reporting no investigations each year does not tally with other available information—the self-reported pressures on researchers to compromise on standards, an increase in the rate of journal articles being retracted, and a growth in image manipulation in articles. Part of the cause may be a lack of understanding of the principles of statistics among researchers, and greater emphasis should be placed on statistical rigour. The sector needs to see increased transparency and reporting of problems as a positive sign that issues are being identified and dealt with accordingly, rather than as a threat.
We see a gap in the UK research integrity system for a new committee to provide a means of independently verifying whether a research institution has followed appropriate processes in investigating misconduct, following similar models in Canada and Australia. The primary responsibility to investigate misconduct should remain with the employer, but there is also a need to improve confidence in the existing system of self-regulation and to adjust for the potential conflict of interest of ‘self-policing’. More broadly, the new committee should be responsible for championing research integrity in the sector, driving the future implementation of a tightened Research Integrity Concordat, and pursuing issues we identify in this report. The new committee will need to be established by and work closely with UK Research and Innovation, and produce an annual report on the state of research integrity in the UK. This is an opportunity for the research community to get ahead of this issue; without such a body being established, there is a risk that the demand for statutory regulation will grow in response to any future scandals, despite a consensus against such regulation within the community.
Meanwhile, there are other steps that can be taken to support research integrity rather than simply responding to problems. We are encouraged to hear that research integrity will form part of the ‘environment’ judgements for the next Research Excellence Framework, and that there are moves towards appropriate publishing of datasets, and better reporting of research methods. Meanwhile, UKRI needs to understand how the pressures and incentives within the research funding system affect research behaviour and consider where counterbalances are needed to ensure a healthy research culture. Training is key to ensuring that the right research culture is imbued by each new generation of researchers and their supervisors, and to ensuring that errors such as common misuses of statistics are avoided. In order to increase the effectiveness of research, increased emphasis should be put on the need to publish ‘negative’ research findings, especially in the field of medicine.
Employers, funders and publishers of research need to be able to share information to support investigations of misconduct, and it is encouraging that protocols are being developed to help employers to manage cases which cross institutional boundaries.
Published: 11 July 2018