Output Records

It is encouraging that all the pilot institutions managed to make submissions of outputs of REF data than had been the case historically. However many of them used the recent RAE data as the basis for the REF pilot and this distorts the actual position somewhat as a huge amount of resource was expended collecting and collating the data for the RAE. All five of the case study institutions relied heavily on outputs already collected for the RAE which were a subset of those outputs actually available and all were clear that some further development will be necessary in order to capture all REF eligible outputs if the final definition is the same as the pilot.

There is widespread agreement that, in order to meet the projected REF requirements, institutions will have to develop much better processes for capturing research outputs. Across the pilot and case study institutions there was a wide variety of models varying from very informal individual academic CVs and departmental catalogues to highly structured institutional repositories. The most common model for holding an institutional catalogue of research outputs is some sort of 'publications database'. The sophistication of institutional publications databases was variable but all were centrally controlled and no institution attempted to maintain a manual centralised record.

Those institutions which did not have any centralised output record, such as The University of Plymouth, were clear that they would need to invest in this area in order to both meet the REF bibliometric requirements and better manage their outputs.