The research assessment process has become a Frankenstein's monster that must be constantly managed in order to minimise its negative effects.
The warning was made at a symposium at the British Educational Research Association (Bera) conference at the University of Warwick last week, which heard tales of "brutality" and "blood on the carpet" from the 2008 research assessment exercise.
However, reporting the results of a Bera/Universities' Council for the Education of Teachers review of the RAE 2008's impact on educational research, Alis Oancea, research Fellow at the University of Oxford's department of education, said that her study had recorded both positive and negative responses.
Many respondents and interviewees feel the process led to their being encouraged and supported in stronger research environments, boosting career prospects, productivity and confidence.
By contrast, others describe a "negative work climate" and poor morale, with deleterious effects on teaching quality, plus concerns that the process narrows the breadth of research publications.
One senior lecturer at a Russell Group university says the RAE forced him "to make compromises in teaching and student support ... that my conscience tells me are unsupportable".
Another says that staff excluded from the RAE had been termed "research-dead" by managers.
In the RAE, education received a relatively low number of 4* ratings in comparison with other social science disciplines, and the study identifies a perception that education departments had lost out as a result.
Tom Schuller, director of the Inquiry into the Future of Lifelong Learning, said that the first two RAEs had had a positive effect, but the process had "become a Frankenstein that was out of control".
John Furlong, a professor in the department of education at Oxford who also worked on the study, said that overall researchers were better off having the process. But he added that academics must "constantly manage this Frankenstein in a way that ameliorates negative impacts".