̽»¨ÊÓƵ

Is University College London really planning to leave half of its staff out of the next research assessment exercise ("UCL memo 'expects'

3*s", December 9)? In the 2001 RAE roughly half the research UCL submitted was graded "national". That equates to 1* in the new system, and would not be submitted under their new policy.

Leaving people out of a submission may move you up a badly constructed league table, but how far up you move depends on how many people other universities leave out. JI've heard a rumour that one institution is going to try to submit only 4* researchers. How many universities can play that game before the next RAE becomes a farce?

How could we prevent such a disaster? The Higher Education Funding Council for England could guarantee that the funding formula will not include thresholds based on the proportion of staff in particular categories - it can still be selective by giving 3* or 4* research more weight. Other funding bodies could also forswear proportional thresholds - if they want to ensure critical mass, they should base this on the absolute amount of high-quality research in a unit.

The Times Higher could promise to calculate its institution-level league table by including all eligible staff and giving zero to those not submitted. Universities would then have a clear incentive to submit all active researchers. We would save all the time and heartache of deciding which people to leave out of selective submissions. And the RAE would give a less distorted picture of the state of the UK's university research.

Richard Green
Birmingham University