The reading test used in the Office for Standards in Education's damning report on standards of literacy in inner-city schools is being redesigned because of doubts about its reliability.
The Neale Reading Analysis, which has been described as the Rolls-Royce of reading tests, is being restandardised just seven years after it was last altered in 1989; until then, the test had remained unchanged since it was devised by Marie Neale almost 40 years ago.
OFSTED's survey, published in May, studied reading standards in three inner-London boroughs - Islington, Southwark and Tower Hamlets - and found that 80 per cent of pupils were not achieving the expected standard in reading and writing. This prompted Education Secretary Gillian Shephard to announce the "national curriculum" for teacher training to ensure that teachers were properly equipped to teach reading.
"Standardisation" involves looking at the reading standard reached by a representative sample of children in order to establish what should be expected at different ages.
It is believed that the 1989 restandardisation used a small sample of children at the lower age-ranges (six and seven), with the result that the reading norms established for these ages were set too high. In other words, the test is too tough for younger children.
This might have produced a gloomier picture of literacy in the three London boroughs of Islington, Southwark and Tower Hamlets than was actually the case.
The Neale test is also used to establish reading ages in all court cases involving special needs support and money for dyslexic pupils. If the reading test norms are set too high, then these pupils could be less behind in reading than they appear.
The National Foundation for Educational Research confirmed that the test is under review and will be republished next year. The restandardisation, said researcher Chris Whetton, would involve a representative sample of 4,000 children.
He denied that there was anything surprising about the decision to adjust the test so soon: "At NFER we have a policy of not allowing tests to become out of date. Reading is something that seems to be changing relatively quickly and also there is the current debate about reading standards."
But Jane Hurry, a researcher in reading at the London Institute of Education, confirmed that there had been doubts among educational psychologists about the 1989 version. "The bottom end of the '89 version is wobbly. There was a big change in the standard younger children were supposed to achieve when you compare it to the previous version. It is a harsh test for the six and seven-year-olds, though it is very fair for older children."
Jane Hurry also referred to a study published in Educational Psychology in Practice in 1994 (Gregory and Gregory, Vol.10), which compared children's scores on the Neale test with their scores on the British Ability Scales test, which is used by educational psychologists.
The reading ages scored by children on the Neale test were as much as 10 months lower than the children tested on the BAS. Jane Hurry added that no children under six were seen when the Neale analysis was being readjusted in 1989, so it was never established what a reading age of five or four would look like.
Dennis Vincent, a researcher in education at the University of East London and the author of two rival reading tests (The Individual Reading Analysis and The New Reading Analysis) also confirmed that there were some questions about the reliability of the Neale test for younger children, adding: "It would be normal to leave things a bit longer than seven years before restandardising a test."