I've been frank in admitting that perhaps the biggest flaw in our old world rankings, which ran from 2004 to 2009, was the "peer-review" element.

A simple reputational survey, it made up 40 per cent of institutions' overall scores. But the sample size was tiny, with only about 3,500 people responding last year.

For the new and improved 2010 Times ÖйúAƬ World University Rankings, we will measure reputation more vigorously. Last week, our rankings data partner Thomson Reuters published details of our approach (see ).

The key features are:

- The survey is run by a professional polling firm, Ipsos MediaCT

ÖйúAƬ

- We are targeting academics who have published in scholarly journals to ensure informed opinions

- The survey is invitation-only

- The poll is issued according to the global distribution of researchers, as measured by the United Nations Educational, Scientific and Cultural Organisation

ÖйúAƬ

- It is in seven languages - Chinese, French, German, Japanese, Portuguese, Spanish and English - to counteract English-language bias

- Scholars will rate institutions in their specific field, allowing them to draw on direct experience

- Respondents can choose from more than 6,000 institutions

- Views on teaching quality will be gathered for the first time

ÖйúAƬ

- Action-based questions, for example, asking teaching-focused respondents to identify the institution they would recommend for the best teaching environment, will elicit more meaningful responses

- The survey instrument will be published in full to allow critical feedback and improvements.

Please take the time to fill it in if you are invited.