The UK¡¯s Research Excellence Framework has been paused for three months to allow for a review of efforts to measure and reward research culture.

Science minister Patrick Vallance has warned against overcomplicating the exercise after the weighting given to people, culture and environment (PCE) was increased from 15 per cent to 25 per cent for the 2029 iteration.

Questions have been raised about whether PCE can be assessed robustly using metrics, with some university leaders warning the credibility of the REF could be undermined.

Here four vice-chancellors give their reaction to ÌÇÐÄVlog.

ÌÇÐÄVlog

Shitij Kapur, vice-chancellor and president of King¡¯s College London

¡°I took comfort from the fact the minister is committed to the REF, committed to having a REF in 2029 and preserving quality-related (QR) research funding. Those three things are very important. We could have heard a minister who said: ¡®You¡¯re struggling to sort out the REF, you don¡¯t need QR¡¯.

¡°Patrick Vallance is one of the few ministers to acknowledge the pressures around full economic costing for research. He didn¡¯t provide a solution, but it shows an understanding of the research pressures that we face and he is responding to these issues.¡±

ÌÇÐÄVlog

On moves to?increase the weighting?of people, culture and environment (PCE) from 15 per cent in REF 2021 to 25 per cent in 2029: ¡°We were not part of the pilot to devise metrics [for PCE] so I haven¡¯t seen the results. However, if the metrics are not robust we have to be cautious about raising the tariff on environment because, in the long run, that will destroy credibility in the REF.

¡°And, if you undermine the REF, you undermine the case for QR and that loss would hugely harm research in the UK, which is vital to ensuring we have talented people in our universities and wider economy.

¡°We would like the REF to be deployed in a way that best incentivises quality and rewards talent in our universities.¡±

David Maguire, vice-chancellor of University of East Anglia

¡°The opportunity to re-evaluate REF 2029 is a good thing. I always favour a research assessment exercise that focused on outputs, not inputs, and preferred the weighting used previously. Any change to the REF introduces more bureaucracy into the system and universities have to adapt to this.

ÌÇÐÄVlog

¡°The last REF wasn¡¯t perfect but it was probably the best exercise that has taken place. Given the last REF was a known quantity we should have not deviated too far from it without good cause.¡±

Nishan Canagarajah, vice-chancellor and president of University of Leicester

¡°We welcome the delay ¨C we should make sure the REF is focusing on quality and impact.

¡°The delay will allow us to clarify the weightings that should be attached to each element of the exercise.

¡°I¡¯m pleased, however, the overall timeline hasn¡¯t changed because universities have already done a huge amount of work on this.¡±

ÌÇÐÄVlog

Rama Thirunamachandran, vice-chancellor and president at Canterbury Christchurch University and a former director for research, innovation and skills at the ÌÇÐÄVlog Funding Council for England, where he developed the 2008 Research Assessment Exercise ¨C a forerunner to the REF

¡°The REF is being asked to do a lot more things, under the current proposals, than it was originally intended to do.

¡°While there are a number of factors that can help to create a good research culture, the REF is not necessarily the right mechanism to achieve this. We should incentivise a whole range of behaviours that should be part of institutional culture, not simply related to research, we need to remain focused on excellence and impact.

ÌÇÐÄVlog

¡°There is a risk that if you add too much to the REF it will become much more bureaucratic and burdensome for everyone involved. Environment should be about whole institutions having a good research culture where they acknowledge and value all staff. It should be about creating an institutional culture which tries to recognise and reward research excellence.¡±

jack.grove@timeshighereducation.com