BU Research Blog

Photo by Shalaka Gamage on Unsplash

BU, like other UK universities, has a support network to help staff prepare for REF 2021. Much of this support centres around three REF categories: Outputs, Impact and Environment. For the past few years I have been chairing the BU REF Outputs Sub-Committee. The committee considers what academics could be doing to maximise their individual outputs and UoAs to maximise the submission of outputs across staff; it also oversees that all outputs are compliant with the requirements for REF submission.

In the latter capacity, the Outputs Sub-Committee oversees the BU Open Access fund to enable staff to publish in Open Access journals that require the payment of a publication fee. One of the key tasks of the committee is to promote REF amongst BU staff, to make sure it is high upon everyone’s agenda, or at least on those members of staff likely to be submitted.

The committee members share ideas and good practice across UoAs. As the different UoAs are of different sizes (in terms of number of staff and hence outputs required) and at different stages of readiness, there is a lot of potential for learning across UoAs. The membership of this committee comprises the Output Champions for all the UoAs to which BU is likely to submit in November 2020. The committee is expertly organised by Shelly Anne Stringer, who makes my life as Chair so much easier.

One issue currently playing is REF2021 changing from using SCOPUS (Elsevier) in REF 2014 as its designated database for ‘checking’ publication data of submissions to Web of Science (Clarivate Analytics). This is particularly important for UoAs in Main Panel A and some in Main Panel B; for BU that is definitely UoA3, UoA4 and UoA11. One would expect that one bibliographic data base is very much like the next one, but nothing is further from the truth.

Knowing that SCOPUS and Web of Science record different outputs, the Outputs Committee approached BU library to investigate. The Academic Liaison Librarian, Caspian Dugdale, took my name as a case study just before Christmas and searched for the various permutations of my name on academic publications. For example, on Scopus there are currently 10 variations on my name. The first finding was that there were 256 publications listed for me on SCOPUS but only 187 on Web of Science. When Caspian compared the two bibliographic datasets, he also discovered that 112 publications were unique to SCOPUS, the larger dataset, but even more interesting perhaps is the finding that 52 records of outputs were unique to Web of Science.

On closer examination, not all records were unique as some simply listed differently on the two databases when attempting to remove duplicate records these were not recognised in the system as duplicates. However, some were unique records, as I had been keeping an eye on my SCOPUS registration since REF 2014 there was nothing new there, but I did pick up two new publications, one from 2013 and one from 2014, that I did not previously know about. Unfortunately, both are half-page conference presentations published in an academic journal long after the conference and both conferences were already listed on my CV.

The main message is that Web of Science appears to be less complete than SCOPUS and that we need to keep a close eye on it to ensure all relevant BU publications are properly recorded.

By Professor Edwin van Teijlingen, Chair of the BU Outputs Sub-Committee

Want to know more?

For more details about how citations data will be used in REF2021, see p.50 of the REF Panel Criteria and Working Methods and p.66 of the REF Guidance on Submissions.

Also, have a look our other BU REF Week blog posts.