The pioneering study of 2,500 children in eight countries reveals that the gulf between Britain and the Pacific Rim countries, well documented in secondary schools, is already in evidence in primary schools.
First findings from the study also show that the difference between the top and bottom scorers is wider in Britain than in any other country and more than twice that in Taiwan and Hong Kong. That suggests that Britain's schools are serving low ability children badly.
Britain scored reasonably well in arithmetic with only Taiwan, Hong Kong and the Netherlands doing better. In using maths only the Netherlands and Norway did worse with the United States, Canada and Australia all outperforming Britain.
Professor David Reynolds of Newcastle University, one of the study's architects, said: "Taiwan and Hong Kong are not cultures for able children. They are doing well because they are serving low ability children well."
The International School Effectiveness project is the first to investigate how culture and society affect pupils' performance and to examine whether what works in Taiwan would work in Britain.
Taiwanese children are taught in classes of 40 or more with the teacher teaching the whole class for most of the time. In Britain most children are in smaller classes and are taught in groups for much of the time.
But the researchers say Taiwanese teaching methods work only because the whole culture is geared to hard work and getting on. Even in the least successful schools with the least competent teachers, children keep on working and achieving.
Professor Reynolds said: "It is very dangerous to say large classes would work in Britain. You can do it if you have very disciplined children and commitment to education. If you have lots of direct instruction as they do in Taiwan it makes no difference whether you have 20 or 40 in a class. But I should be very dubious about the notion that it explains why Taiwan does well."
The project is planned to expand next year to include 12 more countries and to examine which aspects of successful schooling may cross national boundaries.
The Taiwanese and Hong Kong systems have simple aims and, to British eyes, old-fashioned teaching, and they work. But would traditonal teaching methods work here? Would big classes and whole class teaching raise standards among Britain's slowest pupils?
Evidence from the study's earliest findings suggests that they almost certainly would not. The wholesale transfer of apparently successful teaching methods from one country to another is not the answer to educational ills. The reasons for Britain's failure and Hong Kong's success are cultural rather than educational.
Neither large classes nor large amounts of whole class teaching would work here. One of the main reasons British schools switched to group rather than whole class teaching in the Sixties and Seventies was because the latter was not motivating slower pupils. In Taiwan, the work ethic is so strong that children at the bottom of the class keep trying even when they find the task the class has been set is too difficult for them.
Professor Sig Prais of the National Institute for Economic and Social Research said the role of the mother in the culture of Pacific Rim countries was one reason why children there worked so hard. "The mother ensures that a child devotes itself to homework and much more is expected of the role of the home in education."
The study shows that even the way teachers speak to children may have to be adjusted according to culture. In America, the study found positive remarks from teachers - "You are really behaving well today" - led to higher standards. In Taiwan teachers succeed by constantly pointing out to their pupils where they have gone wrong.
Yet, despite the cultural limtis to the notion of good teaching, Taiwan may still have some useful lessons for the world. So may other countries. Can we learn from the Norwegians' success in teaching pupils to be active citizens? Or the Dutch insistence that children with severe learning difficulties should all be in special schools so that teachers face a narrower ability range?
The study suggests that a few characteristics of good schools may cross national boundaries: the amount of time children spend working, close monitoring of homework, excellent feedback to parents and teachers expectations of pupils.
The research's clear message is that all countries need to address these questions because schools affect children's achievement from their earliest years.