A MAJOR Victorian study into how teachers use test data to help children learn has revealed that while struggling students clocked up huge improvements after six months, the performances of top students stagnated.
Lead researcher Professor Patrick Griffin said the ''disturbing'' findings were consistent with NAPLAN and international tests, which also showed results were flatlining.
''It bothers me that our better students are not improving at a level consistent with what their ability might predict,'' said Professor Griffin, an associate dean at Melbourne University.
He said the research had national implications given Prime Minister Julia Gillard wanted Australia to be in the top-five performing education systems in the world by 2025.
''We have a national problem and we have to do something about it.''
The project, which was funded by the Australian Research Council, tested 36,000 students in years 3 to 10 in reading comprehension, maths and critical thinking.
Teachers from more than 500 Catholic and state schools in Victoria were taught to analyse the test results and then divide students into groups depending on their ability, regardless of what year they were in.
A year 3 student who received a high result in maths, for example, might be placed in a group with year 6 students.
''Kids knew if they progressed faster or they struggled they would be moved to a different group,'' Professor Griffin said.
All students were reassessed after six months.
The researchers found teachers were doing a ''brilliant job'' teaching students who scored in the bottom 25 per cent in the first test, with their results six months later showing an improvement rate five or six times higher than anticipated.
Unexpectedly, however, the top 25 per cent of students were, on average, not improving.
''That's what alarmed me - these are a talented, high-ability group of students and we are not developing their potential,'' Professor Griffin said. ''Higher-ability kids should be able to improve at a faster rate than lower-ability kids, but we have the opposite problem.''
He said this puzzled the researchers, who held workshops to discuss strategies for teaching students with different skill levels.
''We were again alarmed, because teachers had oodles of ideas about how to improve kids at the lower levels, but at the higher levels they had fewer ideas.''
Professor Griffin has several hypotheses as to why this is the case. One is that both state and federal governments are preoccupied with ''closing the gap'' between the results of disadvantaged and advantaged students, rather than improving the results of all students.
Another is that some teachers assume high-ability students can learn independently. A third is that the more advanced skills tested are not taught in the curriculum.
A final possibility is that some teachers have found dividing students into ability groups too difficult and have reverted to teaching students according to their year level, which leaves the top students bored and under-extended.
This year the researchers hope to learn more by visiting schools that have been successful in improving the results of high-ability students.
Professor Griffin said the project, which he presented to federal School Education Minister Peter Garrett late last year, could be rolled out nationally and even internationally.
''All the gains in performance are coming at the bottom end and not the top end. If we can get gains at the top and bottom ends, the overall gains will be phenomenal and that will transfer to NAPLAN because we can get kids reading at better levels.''