REF 2028 could revolutionise research culture

The last Research Excellence Framework exercise was a success.

Covid-19 meant that REF 2021 took place in difficult circumstances which impacted everyone involved. Despite this, the exercise was smooth, it commanded the confidence of the sector, and it dealt with a record number of submissions with few hiccups. The aftermath of REF 2021 has been a case study in quiet but honest reflection on some highly technical aspects of research metrics.

REF Now

It would be entirely reasonable to run REF 2028 in the same way as REF 2021. The sector by and large understands the metrics underpinning REF 2021, its language has become part of the vernacular of university staff across the country, and its elements are intuitively familiar to anyone working in research administration. REF 2021 did what it said it would do and delivered more funding to more places than ever.

In reading the results of the Future Research Assessment Programme you get the impression that it is precisely because the mechanisms of REF 2021 were so successful that REF 2028 can be bolder in tackling underlying problems with the research ecosystem.

The 37 or so decisions that inform REF 2028 are in places significant on their own but collectively they represent a significant moment for UK research. This is because REF isn’t just a tool for measuring the quality of research. It is a mechanism for governing the way academics, funders, and universities interact with one another to shape the whole research ecosystem.

The underpinning sentiment of REF 2028 is that individual research brilliance is only made real through effective research environments and the translation of work into impact.

REF Next

Although scoring is the most obvious example of a policy shift toward the collective over the individual it may not be the most impactful. As the starting gun for REF 2028 fires it comes chasing after the Nurse Review, The Independent Review of Research Bureaucracy, and the UK Science and Technology Framework, that all acknowledge that excellent research is a product of the environment in which it takes place. There can be less or more bureaucracy, slower or faster funders, and more or less applied research, but fundamentally research is a team sport and the point of REF 2028 is to expand the definition of what excellence is not just how it is measured.

For example, the proposals should allow more variety in submissions with different forms of evidence. This is both in the way that knowledge and understanding is measured but also in taking greater account of how enabling technologies, cultures, and operations, support research environments.

And it is of course too early to tell how this expanded definition will effectively encourage a greater diversity of REF participants. REF only measures the outputs of the researchers put in front of it after all. However, there is a clear signal toward developing more diverse research teams. This can be seen in the use of a basket of equality measures, the retention of the Equality and Diversity Advisory Panel, and the encouragement for professional service staff to submit as part of REF

In return for broadening the measurement of research there is an emphasis on curbing the worst of the managerial pressures that can come with REF submissions.The flip side of focussing on teams and institutions should also be greater collective responsibility for healthy research assessment cultures.

REF After?

There will undoubtedly be challenges to this more collective approach but for a long time the sector has been saying that it is interested in improving research culture. There is now every incentive to do just this.

Of course, REF does not exist in isolation and there are outstanding questions of how this will fit within wider research and impact measures. In particular with a more environment focus within REF it becomes harder to imagine the unique function of KEF, or how to avoid double-counting data with HEIF. Three measures is likely too many and there will probably be some rationalisation by 2028.

There are also now clearer dividing lines with ARIA which takes the view that research is about getting out of the way of brilliant individuals. Rather than being a competition for which idea will win out there is now the opportunity to build a system with more complementary elements.