Published online by Cambridge University Press: 18 December 2008
This article describes an experiment in the assessment of composition by group consensus. It was designed to prove that small groups of assessors confronted with only aural evidence of particular compositions could, when using suitable criteria, under certain working conditions, come to similar enough judgements on these compositions to prove the method reliable. The method itself was based on the sessions in assessment held at the University of York in 1977 under the chairmanship of Robert Bunting, but, whereas the results in the York session were quite promising and pointed to the reliability of this method, the present experiment has exposed great flaws in a procedure that is advocated by at least one G.C.S.E. group as being suitable for the moderation of students' compositions.