Neil Sears reports on projects which show the benefits of analysing pupils' progress from the earliest years.
The baseline assessments of four-year-olds soon to be required for all new pupils could provide vital information on infant schools' strengths and weaknesses, research reported to the congress suggests.
Studies involving thousands of primary pupils found that progress from the time children enter school until they are seven can vary significantly - not only from school to school, but also for different children or in different subjects within schools.
Research carried out jointly by Harvey Goldstein, of the London Institute of Education, and Hampshire education authority, involved 6,400 pupils in 161 of the county's schools.
The children's baseline tests, done when they entered school at four, were compared with national curriculum tests carried out when they were seven to establish the "value added" by infant schools.
Terry Rath, business and information manager for Hampshire's schools inspectorate, said the most significant discovery made had been the variations between schools. "Some schools have added a great deal to pupils' progress in mathematics, but not for English. And some have high value-added scores in a particular subject for children starting out with low baseline assessments, but are less successful with average children."
The study also found clear evidence that family background affected not just children's starting point but their progress thereafter. Free school meal entitlement was linked to slower progress - not only for individuals, but also for their peers in schools where the percentage of free meals was high. Absences and numbers of primary schools attended were negative factors, though more adults in the classroom and mixed-age classes seemed to give a positive boost.
Peter Coles, Hampshire's director of education, said information provided about effectiveness of schools was proving most useful.
"The value-added analyses are an important additional source of information which helps each primary school to evaluate their impact on school progress in some detail," said Mr Coles.
A further phase in the Hampshire study is focusing on 4,000 pupils from 108 schools whose test scores at seven are being compared with their results at 11.
Pam Sammons and Rebecca Smees, of the University of London Institute of Education, have meanwhile been comparing the reception class screening tests introduced by Surrey in 1993 with national test results when those children reached the age of seven.
Again, schools appeared to differ in their effectiveness: more than a quarter (28 per cent) were performing significantly better or worse than expected on the basis of their pupil intake. In three-quarters of the schools progress was significantly better or worse than might have been expected in reading, writing, maths or science.
As in the Hampshire study, girls made more progress than boys in reading and writing but did less well in maths and science. Older pupils had higher initial attainments on entry and made greater progress.
When these value-added results were fed back to the schools concerned, more than 70 per cent said they found them useful.