ÖйúAƬ

Your recent news story on the research excellence framework quotes Graeme Rosenberg, REF project manager at the ÖйúAƬ Funding Council for England, saying that the proportion of the REF devoted to impact could be as high as 20 to 30 per cent ("Impact may account for 30 per cent of researchers' marks in REF", 23 July).

No doubt there will be a long, intense and unscientific debate about what the figure should be, but no one seems to have remarked on the definition of "impact" being used, especially the apparent exclusion of what is by far the most important impact of university research - on student learning. Without it, what is the justification for conducting research in universities at all?

Nothing could better demonstrate the way in which policies for research and student education continue to be pursued in different silos, the separation of research from teaching, the narrow view of research taken by many researchers, and the power of the research lobby to dominate debate.

Roger Brown, Professor of ÖйúAƬ policy, Liverpool Hope University.