Imagine that a Quality Assurance Agency panel encountered a department that could not tell students whether they would be required to sit a final examination, when that exam might take place, what form it would take and what assessment criteria would be applied.
Imagine that, during the previous assessment period, the examination board had read only half of the material submitted for assessment, had employed no blind marking or second markers, had provided no feedback on individual assessed work and only the most anodyne generalities to groups, and had refused to institute an appeals procedure.
Imagine, finally, that a disproportionate number of starred firsts had gone to the family and friends of the members of the exam board.
Such an analogy far from exhausts the inadequacies of the research assessment exercise ("'Donkey leaders to blame' for RAE failure", THES , April 5), but raises a vital question: even if an assessment regime of some kind is deemed necessary, why should the university system as a whole be subjected to one that would be regarded as grossly inadequate for even the most humble university department?
Reader in history
University of Aberdeen