Today sees the long-awaited publication of the initial decisions by the UK funding bodies for the next Research Assessment Exercise (RAE). The decisions have been based on consultations with UK Higher Education Institutions (HEIs) and the Roberts Review (May 2003).
The Roberts Review
Since the last RAE in 2001, many disputes have taken place over the validity and outcome of the exercise. Responding to this the four UK funding councils commissioned a report from Sir Gareth Roberts, published in May 2003. Roberts’ report recommended that only research departments confident of doing good research should go through the full RAE process, with universities receiving less than 2% of their annual income from funding bodies not bothering with the RAE. In 2002/03 Warwick’s income from funding council grants exceeded 25% of the total income, with further income from research grants and contracts.
The much-discussed proposal that the grading system requires updating forms part of the Roberts review. Roberts recommends abolishing the existing RAE grading system of 1 to 5* and replacing it with "starred quality profiles".
The report also recommended that the RAE continue to be judged by expert peer-review, but allowed for the inclusion of members from industry, business or overseas. This would allow equal emphasis on applied or practice-based research and "blue-skies" research. This also ties in with the conclusions of the Lambert Review on the links between higher education and industry.
Roberts also proposed a two-tier panel structure, with broad subject-panels working with a series of sub-panels (akin to the exisiting RAE panels).
The Roberts Review was sent out to UK HEIs for comment. Through working groups of senior academic staff representing views from all departments within the University, Warwick’s response to the Roberts review was forwarded to the funding bodies, forming part of the consultation leading to today’s RAE 2008 announcement.
Combining the recommendations of the Roberts Review, and the consultation documents produced by around 300 institutions, the UK funding bodies have today published their proposals for the 2008 RAE. The key decisions are:
- The RAE will be carried out ready for completion in 2008
- Like previous exercises, the 2008 RAE will be based upon expert review by discipline-based panels considering written submissions from HEIs. Unlike previous exercises these 70 or so panels will report to 15-20 main panels who will then assess each department. A single assessment method will be used for all participating HEIs.
- As in previous RAEs, HEIs will be asked to identify staff whose research outputs they wish to submit for assessment.
- Due weight will be given to applied research.
- Results will be published in December 2008. The census date (the date determining the affiliation of research staff to a particular institution) is the 31st October 2007, with the closing date for submissions on 30th November 2007.
- To be eligible for submission, research outputs must be published between 1st January 2001 and 31st July 2007. Submissions may list no more than four outputs for each named researcher (although some sub-panels may set a lower maximum).
- RAEs will work on a six-year cycle.
- Results will be published as a continuously graded quality profile for each submission at the sub-panel level (replacing the current 1 to 5* scale). See below for more detail.
- Further consultation on the membership of panels is to be carried out. It is anticipated that panels will include members with experience in commissioning and using research in industry, commerce, and the public sector. It is also anticipated that panels will include members with experience of research in other countries to judge the validity of international research excellence.
A departmental "quality profile" will replace the current grading system of 1 to 5*. The quality profile will have 5 levels: 1, 2, 3 and 4 starred quality levels, and an unclassified level. Departments will be assessed on the percentage of their research submissions that fall into each level. While 4-star will be the highest level, and "unclassified" the lowest, the actual definitions of each level are yet to be decided, and will hopefully be announced later in 2004. The quality profile will show the quality of a department’s research submissions based on proportions, rather than as a single "grade".
To read the RAE 2008 document (RAE 01/2004), which includes a provisional timetable for the 2008 RAE and for more detailed descriptions visit the RAE’s website: http://www.rae.ac.uk