Paul Black and Dylan Wiliam, the King's College London academics at the centre of this year's row over A levels, are used to exam controversies. Both are architects of the national curriculum assessment system, the object of constant sniping by critics.
The two professors' report Standards in Public Examinations, treads on delicate ground, however. While arguing that A-level standards have been "broadly maintained", they query how accurate the exams are at predicting future ability.
Research by Professor Black (left)found that for students who take two similar exams a couple of days apart, the same proportion fail. However, only half of those who failed did so in both exams. He said: "Too much confidence has been placed in the results of external testing and too little in the potential value of school-based assessments."
Professor Black, 71, is a former professor of crystallography at the University of Birmingham. He became increasingly interested in education through his work developing the Nuffield physics curriculum for GCSE and A level. In 1987, he chaired the government's task group on assessment and testing, which devised the assessment structure for the national curriculum. He said: "We stressed that teachers' assessments should be at the heart of any national system. This was accepted in principle and completely ignored in subsequent policy."
Professor Wiliam, 46, is also a former physicist, who studied general science before becoming a secondary school teacher in London. He coordinated the development of national curriculum tests for 14-year-olds in English, maths, science and technology.
He said: "At each (examination) stage, the predictive power for the next stage is pretty weak. It is essential that employers and admissions tutors realise the limitations of exam results and take other evidence of talent into account."