Christopher Frayling, rector of the Royal College of Art, has criticised the handling of design in the 1996 research assessment exercise, writes Kam Patel.
In the RCA's submission to the funding councils' consultation exercise on the RAE, Professor Frayling says that although the definition used by the panel assessing design was sufficiently wide - covering the generation of ideas, performances, images and artefacts including design - "it appears that this was not borne out in practice".
In the RAE, the college was awarded an overall 4, with fine art meriting special commendation as a "flagged research group". Professor Frayling argues that there are many areas of research in the RCA's design schools on a par with or better than fine art.
He says that the problems with design contrast with fine art, where assessment criteria are well established. "In most cases, the end product (in design) is not an exhibition of work plus a catalogue. Instead, research can reach the public domain through a wide, and I suspect to the assessment panel, baffling, variety of means."
Work published in a prestigious professional publication directly parallels an artist's exhibiting in a renowned gallery, the RCA says. But Professor Frayling says that during the RAE "an exhibition in a prestigious gallery was accepted as evidence of research output, whereas a critique in a prestigious professional publication - or the documentation of an important object - were not". He adds that design fields have an accepted practice of using entries to competitions as a way of developing ideas and debates, both technical and cultural. This also was not recognised in the RAE.
THE ROYAL SOCIETY
The RAE is an acceptable way of distributing research funding among universities, but it should concentrate on simplicity, transparency and maintaining stability, says the Royal Society, Britain's most eminent science society.
Its submission says that each assessment panel must contain the right mix of academics and non-academics. Arrangements for 1996 relied too much on the views of outgoing and incoming chairmen, it says. Including overseas members, however, would be impractical. Instead, the society suggests that panel members should have recent overseas experience.
Peer review should remain the main method of assessment, the society says. As the perceived variances between the standards of different panels can create problems, the society suggests a moderating procedure be introduced to help panels adopt a common definition of grades. It says that as definitions are pegged to international standards of performance, differences between disciplines - in terms of the proportion of departments achieving a given grade - are to be expected.
There should not be a requirement of a minimum number of staff for a department to attain the highest grade, the society says, as there is abundant evidence of small departments or sub-departments doing first-rate research.
However, it says, the separate issue of the proportion of staff submitted is more complex. Whether departments entering a small proportion of researchers should be entitled to top grades rests on whether the RAE is seen solely as an aid to funding decisions or whether it also functions provide league tables of research excellence, the society says.
ASSOCIATION OF RESEARCH CENTRES IN THE SOCIAL SCIENCES
The RAE favours mainstream thought within disciplines to the detriment of other approaches, says the Association of Research Centres in the Social Sciences in its submission. In the long term, this will damage the intellectual development of the social sciences and their value to society, says ARCISS, which includes university departments and independent social science research centres.
ARCISS also calls for new arrangements for interdisciplinary research. It suggests three or four more panels to which interdisciplinary research could be submitted and seeks a broader range of expertise on the panels, including research users and overseas researchers.
Like other submissions, ARCISS says the RAE should take place each five years, not four.