The relatively poor performance of US schools in international studies of maths and science attainment can be traced back to unfocused curricula and unsatisfactory textbooks, the conference was told.
William Schmidt of the US National Research Centre said that the textbooks tended to cover too many topics in insufficient depth. They were all too often "a mile wide and an inch deep", he said. As a result, the books encouraged teachers to adopt diffuse learning goals.
Some US maths textbooks, aimed at a specific year group, covered 30 to 35 topics, whereas those in Germany and Japan might average 20 and 10 respectively. In science this trend was even more pronounced.
Mr Schmidt accepted that US maths and science teachers worked hard, often in difficult schools. They might have a 30-period timetable compared with about 20 in Germany and just under 20 in Japan. "The reality, however, is that US teachers are placed in situations in which they cannot do their best," he said.
Schmidt also acknowledged that publishers were not masters of their own fate either, having to face varied, often conflicting, demands from different states and districts. But he suggested that a serious re-examination of curricula was needed.
"In grade 8 science our curriculum is most like that of New Zealand, Iceland, Greece, Bulgaria and the People's Republic of China I we must ask if these are the countries with whom we are, and will be, trying to compete."