HEFCE has today released its Consultation on the Second Research Excellence Framework after a year of delays in light of the Stern Review and now modified from the previous internal draft. In true “hot-take” style up-to-the-minute policy reading, I’ve done a quick first read through of the document and wanted to note some aspects. Note that these are mostly all open to revision in light of feedback to the consultation. The below notes are just the bits that interested/jumped out at me.
- The next REF proposes moving to the submission of all research-active staff, as per Stern. HEFCE propose “research-active” to mean: HESA activity codes of “academic professional” and “research only” or “research and teaching”. There is a measure of independence proposed here to allow the exclusion of research assistants, or otherwise.
- The volume of material to be submitted for each Unit of Assessment (UoA) is to be determined on the number of research-active staff in the unit. HEFCE could prescribe UoA volumes of return by average staff number over census period to avoid gaming/changes to teaching contracts.
- The number of outputs per FTE researcher was suggested by Stern to be reduced to two. There is a worried that this reduces the potential of the exercise to discriminate (para 49). There is a proposal for a maximum number of outputs per single researcher to be six and a minimum to be either zero or one.
- Stern proposed that outputs should be non-portable (i.e. researchers can’t take them to another institution to be submitted). HEFCE notes concern around this, particularly with respect to ECRs and asks for suggestions on how this can practically be implemented. They don’t seem keen in my view.
- HEFCE could still mandate ORCID for REF 2021.
- The move to outputs over a UoE removes the individual staff exemption requirements. There will be an equality and diversity panel. I wonder what this does, though, to departments with a large number of maternity leaves, for instance?
- HEFCE proposes to keep the non-portability of impact.
- Same number of impact case studies as 2014.
- There are new institutional-level impact case studies to showcase interdisciplinary work.
- Proposed quality profile weightings: outputs 65%, institutional impact 5%, impact 15%, environment 7.5%, institutional environment 7.5%.
- The census period for outputs is 2014-01-01 to 2020-12-31.
- Credit for exceeding open access requirements will be based on submitting outputs as OA that are beyond the requirements (eg books) and with licenses and formats that facilitate re-usability. It is not clear what this credit will be, but it is gained by a unit statement on open access.
- The REF after this one (i.e. the one in the mid-2020s, if we live that long) will require monographs/books to be open access. There’s an appendix on this that is very sensitive to all the challenges here, identified in the Crossick report.
I’ll be reading it again and in much more detail in the coming weeks, but nothing unexpected in there in light of Stern and it’s good to see progress and have some information here.