The next REF’s details are more devil-ridden than has been appreciated, says Anna Grey
The general response to the changes recommended for the 2028 Research Excellence Framework, published on 15 June, has been positive. But among those with the closest view of the process, namely REF managers at universities, the reception has been less glowing.
There is, for example, a sense that the rules are changing to accommodate a wider range of output types beyond papers and books. But, in fact, the guidance for the 2014 exercise welcomed any output that met the definition of research. The issue is—and will still be—persuading institutions to submit them.
The increased weighting given to the evaluation of people, culture and environment, up to 25 per cent from 15 per cent in 2021, is predicted to focus university attention on research culture. But with assessment moving towards being based on questionnaires, it isn’t clear how. There is already a vast range of concordats and initiatives in this area—what is the REF going to ask about, and what extra data will be needed?
The biggest change, though, is the decoupling of staff and outputs. The 2021 exercise required each individual submitted to be linked to at least one output and no more than six. Now, the requirement is to submit 2.5 outputs per full-time-equivalent (FTE) staff member, regardless of who produced them.
REF 2021 also introduced a requirement to submit all staff with significant responsibility for research. For most research-intensive institutions, this means everyone on a teaching and research contract.
For others, it means having a process to identify who has time and resources for research. There is no single agreed way to do this; instead, institutions must detail their criteria in codes of practice.
Given all the other rule changes, these codes will probably need rewriting. What will significant responsibility for research even mean, now that any link between individuals and outputs is severed?
Matching exercise
Institutions are expected to comment on the spread of outputs across a discipline and to undertake equality impact assessments on how outputs were selected. This implies that institutions will still need to match outputs to people.
Individuals write outputs, and unless REF panels move to blind review, an element of individual assessment will persist, especially in fields where single-author outputs remain the norm.
This brings us to the next issue. For 2028, anyone employed by an institution for more than six months of the assessment period for at least 0.2 FTE can submit outputs to the REF. Does this really mean any output produced at any point by someone employed at some point in the seven-year assessment window?
The 2016 Stern review advocated non-portability, meaning that outputs submitted had to be produced by people employed at that institution. The new rules suggest a shift to hyper-portability.
At the same time, the requirement for there to be a significant intellectual contribution by each person linked to an output will disappear. This would worry main panels A and B, covering science and medicine.
Another proposal is to accept outputs produced by any member of staff, not just those contractually obliged to do research. Rather than stopping assessment being about individuals, this would drag more people into the REF’s gravitational field and reinforce the idea that the only way to recognise research achievement is through an output in the REF.
If, say, illness or part-time working leave a person without any outputs, will they feel excluded? Does the people, culture and environment section now need to cover all staff?
The REF does not take place in a vacuum. The main complaint of staff is workload, particularly in teaching and administration. They feel that teaching, administration, research support and knowledge exchange are not recognised and rewarded.
Privileging research
By focusing on rewarding those who are able to produce research outputs, efforts to use the REF to improve research culture risk privileging research over other activities, exacerbating problems across the wider university culture.
There is also a danger that research-enabling areas will now seek staff who can also contribute outputs. This could import some of the issues around equity and equality seen in research into support areas and beyond.
Requiring one output for each person submitted drove institutions and departments to support all staff with significant responsibility for research to cross this threshold, while the maximum of six discouraged people from producing huge numbers of papers. Without a link between people and outputs, this might change, heightening competition within universities rather than reducing it.
Anna Grey is director of the research office at Edge Hill University and leads the REF special interest group of the Association of Research Managers and Administrators
A version of this article appeared in Research Fortnight