Exam grades: we should decide how and why we test our children

http://www.guardian.co.uk/theobserver/2012/sep/02/the-big-issue-exam-grades

Version 0 of 1.

Andrew Rawnsley ("Michael Gove has made a cruel mess of exam grades. Discuss", Comment) is right in suggesting that the just solution is "to scrap GCSEs and A-levels altogether and start again with new exams under new names".

The blueprint for such a change already exists – the Tomlinson report of 2004 – which set out plans for end-of-schooling diplomas in which vocational study would be of equal esteem to academic study. Shamefully, the Labour government that commissioned the report also rejected it, fearing the backlash of the rightwing press on the loss of the "gold standard" A-levels. With the school leaving age being raised shortly to 18, now is the time for change.

The recurring concern of the CBI that employers find too many work-place entrants have weak basic skills needs addressing. While anger at the exam boards moving the C grade mark in English between January and June is obviously justified, is it not equally obvious that for a third of GCSE candidates not to achieve a C or above in English is a disaster for them?

It reduces their employment chances, probably limits the extent to which they can enjoy our written cultural heritage and may make for difficulties in handling the bureaucratic demands of modern life.

Michael Bassey, emeritus professor

Newark

Notts

Where exam assessment by grading is concerned, two systems are available: criterion- and norm-referencing. In the former, candidates are awarded their grades by comparing their performance against a set of learning outcomes: 76%+, say, for an A grade, 68%+ a B grade and, depending on the pass mark, 50%+ a C grade. In the latter, the candidates are awarded their grades on the basis of their predetermined ranking within the cohort, the top 15%, for example, gaining an A grade, the next 20% a B grade, the next 30% a C grade, (leading to a 65% pass rate), regardless of their mark.

Although I am not a great lover of norm-referencing it is an effective countermeasure to year-on-year grade inflation. I believe its use is widespread for professional advancement, particularly in the public sector, in continental Europe. If you want promotion, you take a competitive exam set annually knowing that only 20% will succeed. Either way, I don't think it is necessary to follow Andrew Rawnsley's radical suggestion in last Sunday's Observer to "scrap GCSEs and A-levels altogether and start again with new exams".

Dr Stan Morton

Ruthin

Denbighshire

In criticising Michael Gove's attitude towards rising exam grades, Andrew Rawnsley also demonstrates a difficulty in working out what exams are actually for.

It's not true that tests that everyone, or no one, can pass are pointless: otherwise we'd have no driving tests or Olympic 100 metres finals. Are exams there to encourage, motivate or reward?

Or simply to "screen" applicants for higher education, which is tough on the still more than 50% who won't go on to this? Or to drill children for the "world of work", which seems a narrow view of education?

Statisticians will tell you that even accurate measures of one variable cannot be used to assess another, unless the two are perfectly correlated. So you can (possibly) measure an individual child's competence, or knowledge, but not use the same measure to assess how good the school is.

John Old

Nuneaton

Warks

Andrew Rawnsley is wrong to look forward to Michael Gove's retake following the latter's F grade.

If an exam candidate fails so spectacularly and so disastrously as he has done, he ought to be removed from the course and required to find another one, perhaps in waste management, in a different institution altogether.

Professor Colin Richards

Spark Bridge

Cumbria