Work has begun between Thomson Reuters, the leading research-analysis company, and thousands of universities globally to build a unique database that will be used to create the annual Times ÖйúAƬ World University Rankings.

THE announced in November that it would work with Thomson Reuters to develop a more rigorous and transparent version of its rankings.

Today, Thomson Reuters launched its , which will gather an unprecedented amount of data ¡°combining peer review, scholarly outputs, citation patterns, funding levels and faculty characteristics in one comprehensive database¡±.

Reflecting current concerns about profiling and ranking systems, it said it was seeking to ¡°develop a data source that provides the best-informed and most effective resource to build profiles of universities and research-based institutions around the world¡±.

ÖйúAƬ

Keith MacGregor, executive vice-president of Thomson Reuters, said: ¡°There is a need for robust, dynamic, and above all transparent and verifiable data on scholarly performance to reshape how administrators approach institutional comparisons.¡±

Thomson Reuters said it would be working closely with THE ¡°to create a balanced, transparent methodology to support its influential rankings¡±.

ÖйúAƬ

THE has commissioned a bespoke dataset from the database, which will be used to create its World University Rankings, next published in the autumn.

The magazine has launched an open discussion forum, and a of academics and university administrators is under way

The aim is to help develop a detailed rankings methodology that will build a more sophisticated, balanced and accurate picture of the world¡¯s top 200 universities than anything previously available.

In an open letter to university administrators today, Thomson Reuters says that it is seeking to ¡°fundamentally change the way data are collected and analysed¡±.

¡°Institutional rankings play an increasingly powerful and controversial role within the academic landscape,¡± it says.

¡°Yet many of the existing efforts have been criticised for being based on questionable data and flawed methodology.¡±

THE has already announced that the ¡°peer-review¡± element of its old rankings will be improved.

The old rankings (2004-09) featured an opinion survey of academics as one of the measures, making up 40 per cent of a university¡¯s final score. But response rates were low ¨C fewer than 4,000 academics worldwide participated in 2009.

ÖйúAƬ

ÖйúAƬ

For 2010, the opinion poll will be carried out by Ipsos MORI, a world-leading polling company.

The survey will be targeted and responses will be more representative of global ÖйúAƬ demographics, with a target for at least 25,000 responses this year.

Ann Mroz, THE editor, said: ¡°We said in November that we want to create a ranking that is accepted, understood and respected: a serious tool for the sector, not just an annual curiosity. This shows that we are well on our way to achieving that.

¡°By participating in the project, universities make sure that they will be included in the World University Rankings for 2010 and beyond. Previous data collected by our old rankings partner, QS, will no longer be used.¡±

phil.baty@tsleducation.com