Marking shake-up needed to stop exam ‘inflation’, academic says

editorial image

An exams expert has claimed the country should move to a system where a set number of students are awarded top marks each year to avoid rows about grade inflation or the fairness of marking.

Professor Alan Smithers has suggested a “norm referenced” system where a certain percentage of pupils are awarded As, Bs and Cs as a way forward.

The Buckingham University academic was speaking ahead of the A-levels results announcement on Thursday.

Prof Smithers said the increase in students’ A-level results over more than 30 years showed it had not been possible to avoid grade inflation with exams getting easier using a marking system based on a set of criteria.

He said this “inexorable rise” in results had been now halted by Ofqual’s use of “comparable outcomes” in an attempt to ensure standards are maintained. From 1982 to 2011, the proportion of A-level students getting A to E grades increased from 68.2 per cent to 98.2 per cent.

He confirmed since then grades have stabilised by Ofqual, but questioned whether the new system was the best way forward. He suggested it would be better for pupils to be graded based on how they perform compared to the rest of their year group.

Prof Smithers said: “My own view is that this type of manipulation can complicate matters for employers and parents and it would be better to have a system where say the top 10 per cent of pupils get A*s.”

He also said a criterion-based system did not work in practice, adding: “In principle you can specify a pupil gets an A* if they get to a set standard but the problem is that it is too subjective.”

Prof Smithers pointed to the year-on-year increases in A-level results as a sign that criterion- based marking cannot be relied upon to maintain standards.

Under the comparable outcomes approach, Ofqual uses a year group’s performance at primary school to forecast how it is expected the students will perform in later exams. Exam boards which produce results which deviate from the expected grades are then asked to justify this.

The comparable outcomes approach was at the centre of the marking row of GCSE English when several exam boards moved the grade boundaries in 2012 meaning the same standard of work could get different grades depending on whether it was marked in January or June.

It resulted in an unsuccessful legal challenge involving councils, schools and pupils who called for work assessed in June to be remarked using the grade boundaries from January. An alliance of campaigners claimed grade boundaries had been moved unfairly as part of a “statistical fix” to ensure there was no big increase in grades. Critics say Ofqual’s approach makes it impossible for schools to meet rising exam targets because the system is effectively preventing unexpected increases in performance.

But the same system could also protect pupils from sharp drops in grades where they are the first group to sit a new type of exams which might be more difficult or which schools are less familiar with teaching.

Ofqual said: “The approach to awarding we have agreed with exam boards begins with the principle if the group of students taking a qualification in one year is very similar, in terms of ability, to the group in the previous year, then we would expect similar results, all other things being equal.”