Leader: It's a matter of global opinion

When The Times Higher 's first world rankings appeared last year, we acknowledged that they were bound to be controversial and pledged that they would evolve with wider debate. There has been no shortage of that, as university leaders, rank-and-file academics and even an expert group set up by Unesco-Cepes, the organisation's European higher education arm, have grappled with the intricacies of international comparisons. The latest edition, in today's special supplement, takes account of those discussions - the weighting given to peer review has been reduced and the number of academics polled almost doubled, an international employers' survey has been added, the presentation of the ranking simplified and the level of precision reduced to group together universities with similar scores. But the basis of the exercise, combining expert opinion with the few reliable and universally relevant statistics, has not changed. A year of often impassioned debate has not thrown up a radically different approach that works.

Researchers at Shanghai Jiao Tong University, which publishes the main alternative ranking, have also reviewed their methodology and found the same. This summer's comparisons used much the same combination of Nobel prizes and Fields medals, citations and articles as last year's. Their ranking is primarily a test of scientific research strength, in which the age and selection of some data remain the objects of criticism.

In our rankings, the use of peer review has been most controversial.

Academics prefer the certainty of published statistics. But peer review is long established in university life as a method of assessing quality where numerical comparison is not possible. The parallel with university rankings is inescapable: there is no set of statistics that will give an up-to-date picture of the all-round strengths of institutions on different continents.

The best university in the world is a matter of opinion. By asking academics about their own subjects and ensuring a fair representation of disciplines and regions, the outcome should be informed and representative.

Even more than last year, the results show that higher education strength is more widely distributed than the more complacent members of the old guard might wish to admit. Harvard may still be comfortably ahead of the pack, one of seven US universities in the top ten, but European and Asian institutions are well represented in the ranking. UK universities have made significant gains, from the entry of Cambridge and Oxford into the top four, to the sharp rises recorded by Bristol and Nottingham.

In years to come, there will be further improvements. Indicators such as staff qualifications and library spending are possible additions; academic prizes, too, if there is consensus on a range of awards that do justice to different disciplines. The demand for global benchmarking that prompted the rankings will not subside. There remains an open invitation to the academic community to shape their development.

• Go to  The Times Higher World University Rankings 2005