Young researchers feel under pressure to choose topics to suit the perceived preferences of the research assessment exercise, according to a study published this week.
More than a fifth of those questioned agreed that the "perceived research preferences of RAE panels have affected my choice of research topic" in the months after the 1996 exercise.
The figure was significantly greater than the 13 per cent who made the same claim in a similar study that took place after the 1992 RAE.
Ameen Ali Talib, who wrote the report, said: "A concerning trend is the 'gaming' academics are playing and the apparent eroding of academic freedom. An important element of academic freedom is the freedom to do research in topics of one's own choice.
"The more worrying aspect of it is the profile of academics whose research topic is influenced by the RAE panel members' preferences. They are the younger academics in their 30s in lower-rated departments.
"The younger staff also are choosing research topics influenced by the RAE timescale. Some 60 per cent of staff in their 30s agreed that their choice of research topic was influenced by the RAE timescale, whereas the overall figure was 48 per cent."
John Rogers, who manages the RAE on behalf of the funding councils, said:
"If you think about the timing, then, to be submitted to the RAE, the research would have to be done a year before the panel members were announced.
"People often make assumptions that can be significantly off the mark. I have heard before that people stick to more orthodox lines of research but I have never seen it demonstrated."
The study was based on more than 300 responses from staff at English universities.
The work was published as "The continuing behavioural modification of academics since the 1992 research assessment exercise" in the summer 2001 edition of Higher Education Review .