Letter: Survey error

Sir: The article by Lucy Ward "How Dearing's numbers fail to add up"(26 September) far from clarifying the issue of employer support for part-time students, introduced additional erroneous and misleading information.

As the survey was of students rather than their employers, it was not possible to estimate from it the proportion of employers providing financial support, as your report suggests.

It is true that the Policy Studies Institute identified an error in its original estimate of the proportion of employed part-time students whose employers were providing financial support. This was not for the reason erroneously attributed in your article to the author, Dr Claire Callender. It was because students who did not answer a particular question were excluded from the total on which the proportion was calculated, whereas some of them should have been recorded as having no employer support. This meant the proportion with no support was underestimated.

We regret that this misrouting occurred, and that it was not identified in any of the normal internal checks, nor by any of those who read and commented on the report in draft. Such errors do occur from time to time even in major national surveys where several months are devoted to checking the integrity of the data. This survey was undertaken within an extremely tight timetable, which may have contributed to the failure to identify the misrouting.

As soon as Dr Callender realised what had happened she notified the Dearing committee, and contacted the specialist higher education press to enable the information to be placed within the public domain as soon as possible.

PAMELA MEADOWS

Director

Policy Studies Institute

London NW1