COMPUTERS could replace peer review in the research assessment exercise without changing the results, according to a study by Southampton Business School.
Millions of pounds and hours of time spent on the RAE may have been wasted, according to David Watkins, professor of management development and Patrick Thomas, lecturer in management information systems.
They reconstructed elements of the last exercise in management studies, ignoring peer review findings and using citation data alone. The final assessment turned out to be very similar. The method relies on citation data banks based at the Institute for Scientific Information in Philadelphia.
These record each time a work is cited in a refereed journal, and give it a weighting based on the journal's international importance. Professor Watkins said: "Under the Dearing proposals of eliminating the bottom half, we could virtually automate the whole process, saving a lot of time, money and turmoil."
The Southampton study was able to fine-tune these values by ranking works according to their impact within a single discipline.