ÁñÁ«ÊÓƵ

The argument presented by Simon Szreter on A-level grades (THES, November 17) could hardly be more misconceived. Data collected by the ALIS survey, run by Carol Fitzgibbon, professor in the University of Newcastle's education department for many years, demonstrates that candidates of comparable quality (determined on a numerical basis by average across-the-board scores at GCSE) achieved far lower A-level grades in mathematics and natural sciences than in either humanities or social sciences.

This is purely objective data, based on thousands of candidates from hundreds of schools and colleges, and without prejudice to the interests of any particular subject area. The results of the survey show that physics candidates are the most disadvantaged of all in this respect, followed by those in chemistry, maths, general studies and biology. Those in Dr Szreter's own subject come way down the list, after economics and three modern languages.

The predicted A-level result for a physics candidate with a full complement of As at GCSE is considerably less than a certain B. In the case of physics, my own investigations have shown that this is most probably due to the progressive elimination of the less able candidates from the subject, partly due to the increasing pressure on institutions to produce "results". The desire to maintain approximate percentages of particular grades in the current norm-referencing procedure has, by a positive feedback effect, led to a situation in which competition between increasingly higher-quality candidates at each level is gradually pushing up the standards required, while the continual migration of candidates to subjects perceived as less difficult pushes down the standards required in those areas.

The result in the case of physics has been catastrophic - a drastic fall in numbers. The reason for the problem, however, is becoming widely recognised by such organisations as the Institute of Physics and by the recent meeting of the National Association of Head Teachers in Dublin. It is unfortunate that Dr Szreter (and, one suspects, other people in the university sector) should be unaware of the vast amount of work that has been done, and that his proposed "solution" to the problem of the imbalance of grades at A level would help to perpetuate its cause.

Peter Rowlands De La Salle College, Pendleton, and University of Liverpool