National tests too imprecise to measure value added
It seems that infant schools that want to show good results at seven may take advantage of imprecise rules in the administration of key stage 1 Sats, which means their seven-year-olds appear to do better in Year 2 than children from all-through primaries. Primary schools need to show good progress between seven and 11, so they would be disadvantaged if they inflated their Y2 results.
The study, commissioned by the Department for Education and Skills'
innovation unit to look at the transition between key stages 1 and 2, found that pupils' scores on KS1 tests drop sharply between the end of Y2 and the start of Y3.
Almost one in five of the 339 children in the study gained the top grade, level 3, when they sat the writing test at the end of KS1, but when the test was repeated the following autumn, none of them reached that level.
There was also a sharp drop in the proportion achieving level 2. Scores in reading and maths also fell between the end of Y2 and the start of Y3. The difference in test scores was smaller in the two primaries than in the six junior schools. But the results improved dramatically after two terms in Y3.
John Pugh, head of Maryport junior, which took part in the research, said:
"There is probably more pressure on people in infant schools to get higher results."
This pressure on infant schools to give borderline pupils the benefit of the doubt means that baseline results used to calculate the value added by junior schools were higher than they should be, he said.
Infant schools were more likely to take advantage of rules allowing teachers to give pupils more time and adult support.
Mr Pugh said tests at the start of Y3 would provide a more accurate measure of the value added during KS2.
The report, "Transitional Regression of Primary School Pupils", says more precise tests than the Sats would be needed if the results were to be considered useful to KS2 teachers. "It would appear, from the results of the tests given, that regression due to holidays, new schools and lack of practice on test material, could not have resulted in such a dramatic regression in attainment", the report says.
"It would seem to be the process of testing which is flawed as there is an opportunity for inconsistency and the interpretation and administration of test material by different schools and different teachers within schools, which could lead to inaccuracies."