Points for Institutions to Note

1. Without adequate ICT support, compiling the REF pilot data was a labour intensive and time consuming process. Given that the pilot came immediately after an RAE exercise, and some institutions were able to re-use RAE data, future manual production of REFs will be even more challenging
2. HR systems and processes need minor amendments to capture previous and destination institutions if these are required. Processes for tracking early career researchers will need to be built into both HR and other research management systems
3. It is probable that in the medium term institutions will wish to hold all 'table two' outputs in some form of 'outputs catalogue'. Institutions should seriously consider implementing at least a publications database and view this in the wider context of their overall research management and the strategy for promulgation of research results. If appropriate an institutional repository should be considered
4. The strategy for populating a publications database or institutional repository should be carefully developed and widely consulted upon. Due note must be taken of the fact that the final specification of the REF will comprise more elements than the bibliometrics/citation analysis. Strategies and systems will need to be put in place that support both the bibliometrics and their relationship with WoS and SCOPUS as well as cataloguing other research outputs that will form part of the peer review and community components
5. Production of the link table between the research outputs and staff tables can be very challenging if carried out manually. There are two possible strategies to help in this:
   1. Institutions should consider the degree of integration possible between their HR and publications systems. It is probable that explicit links between the HR system and publications catalogue will make preparing a REF return easier but will present maintenance challenges, particularly in cases of multiple authorship.
   2. Institutions should carry out a cost/benefit exercise to determine whether there is a case for developing comprehensive research data warehouse which brings together all aspects of research data. In case study institutions that had one, such a warehouse made the production of the link table a simple reporting process. However institutions should also note that the final REF specification is not yet defined and the data needed for the pilot is likely to be a superset of the final specification.

Definitions of Staff Categories

1. Category A: academic staff in post and on the payroll of the submitting institution on the census date. Eligible Category A academic staff must be employed under a contract of employment with the HEI on the census date. Their contract must list research and/or teaching as their primary function

2. Category B: academic staff who held a contract with the institution after 1 January 2001 and who left the institution (or transferred into a department returned to a different UOA) after that date and before the census date, and who otherwise would have been eligible for inclusion as Category A

3. Category C: independent investigators active in research who do not meet the definition for Category A staff but whose research on the census date is clearly and demonstrably focussed in the department that returns them

4. Category D: independent investigators who met the definition for Category C staff during the period 1 January 2001 to 31 October 2007 but not on the census date

More information can be found in the Guidance for Submissions