RAE loophole closed but confusion remains

Funding chiefs have moved to close a loophole that could allow game-playing in the next research assessment exercise, scheduled for 2008, by continuing to exclude less-prestigious researchers.

But academics remain confused about the extent to which the new quality profiles - under which the proportion of work produced by a department will be graded from 1* to 4* - will result in the omission of weaker work.

Marcus Doel, professor of geography at the University of Swansea, who brought the issue to The Times Higher 's attention, said: "If a group of staff each have one piece of 4*-rated research and three pieces of unclassified research, they can submit either one piece and get 100 per cent top-rated (resulting in a perfect quality profile) or four pieces and get 25 per cent top-rated and 75 per cent unclassified (resulting in a disastrous quality profile and probably rapid closure).

"The logic of this zero-sum game is that wherever possible institutions should submit all researchers (for the multiplier) but only one piece of their research, which is judged most highly on the pre-defined standards (to maximise proportions)."

But the Higher Education Funding Council for England has told The Times Higher that departments seeking to do well should submit the maximum amount of work per person assessed as defined by the panels that will judge it.

The advice points to the continued exclusion of some researchers from the exercise.

A spokesman said: "Based on our experience in previous exercises, we would anticipate that where a panel asks higher education institutions to submit up to four cited outputs for each researcher in a submission, the full number will be cited in the great majority of cases.

"While it remains our intention that panels should not unreasonably mark down a department in which fewer than four outputs were listed for some staff, there is a point beyond which a department that cited significantly fewer without good reason should expect the panel to view this as evidence of a lack of depth in its research activity overall, to be reflected in the rating or profile awarded."