The study claims splitting children into ability groups does not raise standards and depresses the performance of working-class pupils.

The four-year study of setting in maths from key stage 3 to GCSE found that it helped pupils in the top sets to make progress at the expense of those in lower sets, who slipped further back.

Pupils made most progress at the schools that waited till Year 10 to introduce setting . But these schools also used little whole-class teaching, continuing to use a wide range of methods, including groups within class, suggesting reasons other than setting may be a factor in their performance.The pernicious effects of setting were far greater in the four schools that used formal whole-class teaching, the researchers said.

The study of six schools, by Dylan Wiliam and Hannah Bartholomew, of King’s College, London, found the set in which students were placed made far more difference to progress than the school they attended.

On average, being placed in a top set would raise a student’s maths GCSE score by more than half a grade; being placed in a bottom set would lower it by the same amount.

The researchers, who presented their findings this week to the BERA conference in Leeds, also found that setting reinforced class and gender differences in achievement.

At two of the schools, working-class students were placed in lower sets than would be expected from their key stage 3 scores, which depressed their achievement.

And higher-attaining boys in bottom sets achieved up to a whole grade less at maths GCSE than girls who had similar scores in their KS3 tests.

Email: dylan.wiliam@kcl.ac.uk