Research contest off till 2001

THE research assessment exercise may be delayed by a year to 2001.

Submissions to the funding councils' consultation on the exercise - the method by which the councils assess and selectively fund research in universities - must be in by today. Early feedback from regional consultation meetings suggests that a fifth RAE is on the cards with points of detail, rather than the principle of the RAE, the main stumbling blocks. The exercise has been carried out every three or four years since 1986.

A joint submission from the research-led Russell Group and the 94 Group of universities calls the exercise's fundamental approach "sound", but says "doubtless details can be improved". A submission by the Council of Modern Universities is believed to echo this view.

Sources close to the funding council say the areas raising most concern are the timing of the next RAE, the comparability between panels, the spectre of a transfer market in strong researchers, and whether departments that enter only a small proportion of their researchers should be able to score top grades. The make-up and chairing of assessment panels is also a bone of contention.

The Russell/94 groups' submission says that an "equilibrium" is being established, with only slight changes in ranking between each of the RAEs. "This fact, combined with the need for a generally lighter touch, leads us to recommend that the next exercise be held in 2001," it says. The submission rejects the idea of a rolling programme.

The transfer market, where top researchers are scooped by universities in the run-up to the RAE, is widely criticised. The Russell and 94 groups say it has "distorted the academic labour market in a minor but significant fashion". Initial indications from the regional meetings are that feelings are sufficiently strong to force change. In future the RAE may reward both the department where the research was undertaken and the department where the researcher is now.

Submissions want comparability between panels, as well as transparency and impartiality, with some support for the idea of an advisory super panel, whose members could attend panel meetings as observers. This way universities could have confidence that a 5* in one subject compares with a similar grade in another.

The make-up of panels has also been raised, with concern that the same people sitting on panels in successive RAEs may lead to excessive conservatism. The Council of Modern Universities wants greater representation on the panels and says the outgoing panel should not appoint the chair of the new panel.

Other submissions stress the need to appoint the best researchers regardless of institutional affiliation and previous RAE service.

Feelings run high over whether departments should be credited with top ratings if only a small proportion of staff enter the RAE. A composite score, reflecting quality and proportion of staff submitted, is a possibility.

The government raised the prospect of drawing on research expertise from outside the United Kingdom in its response to Dearing. Submissions propose international panels that could provide postal references when requested.

Quantitative measures of assessment, such as citation indices, are likely to be re-examined, but the Russell/94 groups argue that peer review "remains by far the most reliable method of assessing research quality".

Calls to relax the definition of research have also been voiced, particularly by the modern universities. Support for formal feedback to departments after RAEs appears to be slight.

PROVISIONAL TIMETABLE FOR NEXT RESEARCH ASSESSMENT EXERCISE

March:

Written submissions to the RAE consultation due today.

May:

The funding councils must present to their boards an analysis of the submissions. Recommendations on the shape ofthe next exerciseare also expectedto be presented to the councils.These are likely to include viewson panel make-up, ideas on how to address the problem of the transfermarket and viewson whether thesubmission date should be brought forward.

July:

The fundingcouncils will issuea second consultation paper.

Autumn:

Responses to the second consultationshould be in.

January 1999:

Firm decisions on the next RAE published.

What may be harder for the funding councils to provide in advance of the next RAE is details of the funding framework that will accompany it. This is despite a call from the Russell/ 94 groups for the funding framework to be in the public domain at least three months before the submission date.