Professors Peter Mortimore and Harvey Goldstein (Guardian Education, February 25) are unhappy with the data in my reportStandards of Reading, Spelling and Mathematics for 7 year olds in Primary Schools in 1995 and critical of my policy conclusions. I will come to their criticisms later. But first let us set the scene.
15 years ago, my colleagues and I tried to conduct a census of the examination results of all the secondary schools in the country. But we failed because many schools would not send us their results, which legally they had to publish, unless we promised them anonymity. So we settled for a representative sample of over 2000 schools whose identities remained secret.
Our most striking finding was that good results - so vital to children's future careers - varied dramatically from school to school, even between schools of the same type in similar areas. Factors of four between the top and bottom comprehensive schools were the norm all over the country - with even greater differences in inner city areas.
We suspected then that such enormous differences did not start in secondary schools. But at that time detailed data on standards in primary schools across the country was hard to find and even harder to acquire.
Now, thanks to the National Curriculum and the national tests which go with it, we do have much more detailed information about standards amongst 7 and 11 year olds.
So when this data for 1995 - for most of the nearly 19,000 English primary schools - was made available for analysis, I applied to the Curriculum and Assessment Division at the DfEE (Room 4.10, Sanctuary Buildings, Gt Smith St, London, SW1P 3BT) and, a month or two later, got down to work.
Much the most striking findings are the very large differences in standards between schools both across the country and within the same Local Education Authority (LEA). Pupils in the top quarter of schools are about nine months ahead of those in the bottom quarter in Reading, Spelling and Maths. And within the same LEA, pupils at the top schools are, on average nearly three years ahead in these basic subjects compared to those at the bottom schools. And all this after only two years of compulsory schooling.
The relative differences are even greater than for 11 year olds where my earlier study had shown that - after six years of compulsory schooling - pupils in the top quarter of schools are more than a year ahead in English and over 18 months ahead in Maths of those in the bottom quarter. And within the same LEA, pupils at the top schools are, on average nearly four years ahead in English and five and a half years ahead in Maths compared to those at the bottom schools.
Professors Mortimore and Goldstein object that the relationship between National Curriculum levels and years of pupil progress is not perfect and that therefore we need much more sensitive measures of pupil progress such as the standardised tests already used by so many schools.
I quite agree. I spent most of 1993 arguing for this, winning the argument but losing the policy battle. And I upset Sir Ron Dearing and others by publishing, in February 1994, an earlier report for the Social Market Foundation - Why there is no time to teach: What is wrong with the National Curriculum 10-level scale - which argues the case in detail.
But National Curriculum levels are the only data we have for schools across the country. So the alternative is not to use something better - it doesn't exist - but to sit back and do nothing about the massive levels of underachievement which the test results reveal for the first time.
The two professors also rebuke me for not doing something else - such as a value added analysis - when the required data are not available and will not be for many years.
So could I suggest that people actually read the report - Standards of English & Maths in Primary Schools for 1995 (both from the Social Market Foundation, 20 Queen Anne's Gate, London SW1H 9AA, price £10) and make up their own minds about whether something more should be done - now - to give many more of our children the start they need or whether we should wait until another generation or two of schoolchildren have underachieved as badly as the current generation
Better still, why not apply to the DfEE for the data and check the results for yourselves. There is surely much more that this - the richest database ever collected about English primary schools - can tell us.
And even if there are some problems with parts of the data, we surely cannot ignore the main message - that there are severe problems in the teaching of Readng and Mathematics in very many primary schools across the country.
When the report was published, both Gillian Shephard and David Blunkett agreed that urgent action was needed to try to share, in David Blunkett's words, "best practice from succeeding schools with those who are struggling with unacceptably low standards of achievement."
The Government should also be supported by all in education in its policy of publishing - school by school - National Curriculum test results for 11 year olds. And, in future, such performance tables should include precise scores for reading and arithmetic and should also be published for 7 year olds.
This would start, for primary schools, the long process of public discussion and changes in practice which have gradually been bearing fruit in secondary schools in the five years since the Government started publishing performance tables for GCSE and A-level results in 1992.
All those in education should be challenged to support such a policy of open public accountability. For me - and I suspect for many and especially parents - this is the acid Nolan test for all those professionally involved in education.
Are they for or against such publication? And if they are for publication, are they prepared to say so in public? Or is their motto "Not in front of the parents."
And this is a test which, as far as I am aware, neither Professor Mortimore nor Professor Goldstein would pass.