Paul Newton's work on the definition of exam standards (TES, July 8) is seductive but naive. He claims that there would be greater clarity if a single, all-purpose technical definition could be found to underpin exam standards from year to year.
It is precisely because educational standards mean different things to different people that they are such fertile grounds for debate. To hide behind a single chosen statistical technique and rule any challenges to it out of court by definition is simply to take bureaucratic refuge from educational complexity.
Instead, we award grades in a way which reflects the complex nature of standards. But we also need to be accountable and willing to explain our standards.
To this end, AQA's results this summer will be accompanied by a booklet which explains exactly what evidence we use, and how we use it, to set the standards which give all AQA's candidates the grades which they deserve.
Mike Cresswell Director general Assessment and Qualifications Alliance Stag Hill House Guildford, Surrey