The nine-country report, which is due to be published early next year, will show that the basic and higher-order maths skills of English, American and Dutch pupils are markedly inferior to those of Taiwanese and Hong Kong children. It will also reveal that English nine-year-olds are, on average, 18 months behind their Taiwanese peers.

The International School Effectiveness Research Project team acknowledge that their findings are open to challenge because the school samples were small - between 5 and 12 primaries were visited in each country - and they included an unrepresentatively high proportion of very effective and less effective schools. Nevertheless, the research will cause concern because it also revealed that England was the only country where the span of maths attainment widened rather than narrowed during the two years that the schools were being monitored.

Furthermore, the study found that England was one of only two countries - the other was the United States - where schools were of very variable quality. In Taiwan and the Netherlands, on the other hand, the differences between schools’ maths scores could be mostly explained by referring to children’s IQ scores and families.

Another worrying finding was that the quality of teaching observed in Taiwan, Norway and the Netherlands was significantly higher than in England, the United States and Hong Kong. The colony’s teachers will no doubt dispute that, however, as their pupils finished top of the basic skills table (the Taiwanese were second in this table but ranked first in the higher-order skills section).

Professor David Reynolds of Newcastle University, one of the project’s leaders, said that the findings raised fresh doubts about British education. “The study is not yet fully analysed but one must be concerned that our findings are so similar to those of other researchers in showing a relatively poor British performance,” he said.

“Later this year we hope to explain what it is that is going wrong. The evidence would, however, make one suspicious of the British practice of reducing the amount of direct teacher input to a small proportion of the lesson time and the related practice of differentiation of pupils, both of which may serve to widen the range and reduce the average levels of performance.”

Two other international studies have drawn attention to the poor maths performance of British children. In March, a 17-country study by Exeter University researchers revealed that English and Scots secondary pupils started from a lower base and made less progress than pupils in other countries. And in January the National Institute for Economic and Social Research reported that British 10-year-olds were as much as two years behind their continental counterparts.