Assessing student performance in examinations based on the content of lectures should be straightforward, but if universities adjust marks when things go wrong, a lack of transparency can leave students feeling less than satisfied ("Litigation fees top ?15m as academic disputes grow", March 12).

My university decided to adjust upwards the grades of a large group of students who had performed poorly because of failings in the conduct of a physics module. When I queried the methodology, because I was concerned that the adjusted grades were not a fair representation of students' abilities, I hit a wall of silence.

First, the school argued that the adjustment was not made on a statistical basis because the sample-size group was too small to make fair statistical comparisons with previous years' grades. Then, perhaps realising that the implication therefore was that results had been treated arbitrarily, the university refused to discuss the matter further.

In a rapidly expanding university sector, poor teaching, poor lecture planning and poor management are bound to occur in isolated pockets. If students pay a contribution to their courses, they are likely to complain about such bad practice. If treated sensibly, complaints can have the positive effect of raising standards.

ÖйúAƬ

Where mistakes have occurred, the problems should be raised in an open and transparent manner. This helps to placate disgruntled students and reduces the threat of them seeking legal redress.

When modules or subject areas have not been taught in accordance with syllabuses, and students fail en masse, universities should acknowledge their mistakes and discount grades across the board, rather than attempt a seemingly arbitrary adjustment.

ÖйúAƬ

When examiners try to adjust grades for large groups of students, but fail to provide evidence of a rigorous (or any) method for their adjustment, students question the method and validity of their assessment.

Have readers experienced similar problems with "adjustment" procedures or can they offer examples of best practice?

Owen Wainhouse
Cardiff