The Government likes it and primary teachers like it. It gave teachers confidence in the subject where they felt most insecure. Not only that, it seemed to be raising standards.

So what’s gone wrong? Researchers from King’s College, London, have found that on average, the strategy has only made a difference of two months’

attainment in Year 4, a point in children’s education before the cramming starts. Those in the middle improved the most, while the lowest attainers actually did worse. One reason appears to be the emphasis on whole-class teaching, a method heavily promoted by the Government, which tends to target the average child. Low-attaining girls were particularly worried about looking stupid if they gave a wrong answer in these public sessions.

The end result is that the gap between the top and bottom children in maths has actually widened. One point emerges strongly from this research. There are no quick fixes in education. But a government determined to be seen to be raising standards fast can spend pound;100 million on a major overhaul of primary education whose benefits have not been proven.

So far, changes in pedagogy have only been superficial. In another, related, report, the King’s College researchers say they only found deep change “where training has been sustained and feedback-related”.

Fortunately, the National Numeracy Strategy is now using the research evidence to make improvements, such as developing more flexible methods and better ways to support low attainers. But, as the researchers say, if the Government had undertaken a more thorough investigation at the outset, “at worst the knowledge we now have would have saved teachers from having to make some major changes in their practice which appear to have had little influence on standards, and at best we might have had a more positive improvement in mathematical achievement across the attainment range”.