ÖйúAƬ

The major miscalculation in relating student-to-staff ratios to time spent teaching is the use of percentages of time instead of hours ("The lecturers who don't teach (but still count)", News, 6 December). This error is deliberately required by the ÖйúAƬ Funding Council for England and is also used by Research Councils UK. Universities are not expected to report the number of hours per week worked, only percentages. When RCUK provides funding for 10 per cent of an academic's time, this is stated to be 37.5/10 hours, or less than four hours. However, there is wide variation between and within institutions in the number of hours worked per week. My typical week would be at least 50 hours, both in term time and during vacations. Sixty hours is not uncommon in many disciplines.

It is illogical and incoherent to compare percentages based on different working weeks. For a 37.5-hour week, 50 per cent is about 19 hours; for a 60-hour week, 25 per cent is 15 hours. The difference is much less dramatic than the percentages suggest.

The other serious problem in comparisons at the university level is differing disciplinary mixes. Laboratory subjects necessarily have more contact hours: students are allowed to work in libraries without supervision, but not in labs.

Jane Hutton, Department of statistics, University of Warwick