Social scientists are divided over possible use of citations in the REF

Social scientists are divided over whether "bibliometric" data, such as the number of times research papers are cited by peers, has any place in the new system being designed to replace the research assessment exercise.

The UK university funding councils have said that under the system to replace the RAE - the research excellence framework - research quality in social sciences, arts and humanities will be judged by an unspecified "light-touch peer review", starting in 2013.

This could include the use of some bibliometric data such as citation counts to inform peer review.

In contrast, science-based subjects will be assessed using three metrics indicators - research income, postgraduate student numbers and research citation counts - phased in gradually from 2010.

At a meeting organised by the Academy of Social Sciences in London this week to discuss its position on the REF proposals, social scientists heard that the quality of their research could easily be judged by citations counts - as long as books and policy publications were included in the data.

Sylvia Walby, professor of sociology at Lancaster University, has been examining bibliometrics. She said that in most social science subjects less than half of the scholarly communication took place through journal articles. But if the sources were broadened to include books and policy reports, citation analysis could help reduce the amount of material peer reviewers had to wade through.

"It is quite possible to utilise bibliometrics for the social sciences as one of the measures in a light-touch peer review if we include books and policy publications. Can it be done? We already find (the data we need) in both Google Scholar and Thompson's Web of Science," she said.

But other academics questioned the wisdom of using any bibliometrics in the social sciences at all.

Susanne Kord, head of the department of German at University College London, said: "You will get more citations if you work in a growing field than a shrinking field ... the inclusion of metrics will influence people's decisions on their choice of research subject, and they will come under pressure from the university leadership to focus on the growing fields.

"The fields in which real discoveries can be made will fall by the wayside."

Addressing the meeting, Graeme Ronsenberg, an officer of the Higher Education Funding Council for England, said the funding council was not making any proposals about how bibliometrics could inform light-touch peer review for non- science subjects at this stage because it was busy concentrating on the sciences, which would be assessed earlier. Work on the light-touch peer review will start in 2009.

"By moving ahead quite quickly with the sciences and leaving the serious development of light-touch review until a bit later, we do realise this creates some unease," he said.

zoe.corbyn@tsleducation.com.