Out for the count

Your report on graduate employment perhaps understandably does not challenge the Higher Education Statistics Agency data but simply reiterates the published so-called performance indicators ("Thames Valley graduates on the right track to work", 17 July).

Hesa requires that the data collection must have a minimum response rate of 80 per cent to be acceptable and yet half of your "top ten" had a response rate of less than this. One was based on less than two thirds of the student population, so the conclusion that the indicator was more than 90 per cent is based on the assumption that more than one third of the students who did not respond had the same level of employment statistics as those who did. Furthermore, often 15 to 20 per cent of the respondents are dismissed from the base population because they are "not available for work" or "refused to take part".

The result is that these so-called performance indicators are commonly based on less than two thirds of the eligible student respondents. If there is less than an 80 per cent response rate Hesa should not publish the data.

It is also worth noticing that massage therapists and dressmakers are considered graduate jobs whereas air-traffic controllers and transport managers are non-graduate.

Mike Page, Deputy vice-chancellor, University of Huddersfield.