As test scores climb, sceptics say papers are getting easier. A new study should give them pause for thought, writes David Budge.
THE improved national curriculum test results for 11-year-olds announced this week reflect a real improvement in standards, new research suggests. Despite claims to the contrary, neither the papers nor the marking appear to be getting easier.
A study comparing children's performance in the key stage 2 maths test and a second test devised by the National Foundation for Educational Research - - called Mathematics 11 - has found that the level of difficulty of national curriculum tests has remained consistent over the past four years. Initial evaluation of this year's maths test also suggests it is no easier, says Dr Mike Treadaway, of the Fischer Family Trust, which funds research into education.
Dr Treadaway, a former adviser in the Vale of Glamorgan, found that the marking in English had also been consistent - except in 1995 when the grades awarded were "about half a level too low". His study also suggests that teacher assessments - particularly for English at key stage 3 - are a marginally better indication of pupils' future attainment than test scores. But he says that the most reliable analyses are obtained when both test results and teacher-assessments are used.
Dr Treadaway has analysed data from 19 Welsh local authorities, with the aid of fundin from the Welsh Assembly. He has also studied data from England that provide information on a further 500,000 pupils a year.
"This week the usual arguments about the validity of national curriculum tests are, once again, receiving publicity," he says. "But we should look at the evidence rather than hearsay.
"Overall, national curriculum levels are just as accurate as other tests in providing an estimate of potential attainment."
Dr Treadaway acknowledges that many secondary staff do not believe that the key stage 2 scores are a reliable indicator of children's ability.
However, he says the test scores are a slightly better predictor of key stage 3 results than the CATs (cognitive aptitude tests) scores for Year 7 pupils.
"Whilst there may be some differences between primary schools in terms of the reliability of key stage 2 data, these are no greater than the variations in Year 7 test scores, which possibly reflect how well pupils settle into secondary school.
"In future, it will be possible to see if the relationship between, for example, key stage 2 and 3 attainment remains consistent. Will the improvements in 1999 at key stage 2 be reflected in the 2002 key stage 3 results? Much might depend upon whether secondary teachers regard the levels achieved in 1999 as a basis upon which to build."
Contact: Dr Mike Treadaway on e-mail: email@example.com