For more than 15 years Dr Marks has retained a high media profile by attacking peace studies, the National Union of Teachers, declining GCSE and A-level standards, vocational qualifications and HM Inspectorate. He has been extremely influential but has often been accused of drawing sweeping conclusions from inadequate data. He once confidently declared that whole-class teaching is about four times as effective as teaching children in groups. And statistics he produced to prove that there had been a slippage of standards in inner London secondary schools were dismissed as "buffoonery" by the ILEA's then education officer, Peter Newsam.
The same cannot be said about the study of key stage 2 maths and English results that he has produced for the Social Market Foundation. It is true that the report offers too much information that we knew already. Girls are outperforming boys, independent school pupils are doing better than their state school peers, and the majority of Year 6 children are failing to reach level 4, the standard expected of 11-year-olds.
But at least three of his other findings will cause shock waves. Within the same LEA, pupils at the "best schools" are on average nearly four years ahead in English and five-and-a-half years ahead in maths compared to those in the "worst schools". Ninety per cent of pupils in the "best school" in an LEA reach level 4 in English whereas only 10 per cent do in the "worst". And pupils in the best-performing authority are on average more than two years ahead in maths compared with those in the bottom LEA.
These are depressing findings and well-rehearsed arguments about large classes and an overloaded curriculum are hardly an adequate defence. But once again Dr Marks has weakened his case by basing his conclusions on unsatisfactory data.
He fails to acknowledge that there was general agreement that last year's English and maths tests - the first to be carried out nationally - were too hard. Even 11-year-old maths prodigies who obtained a C grade at GCSE (equivalent to level 7 or 8) last year only managed a level 5 in the key stage 2 test. On the other hand, the science test, in which 70 per cent of children reached level 4, was too easy. Dr Marks leaves the science results out of his calculations and offers no explanation.
It would also have been fairer to point out that some schools were accused of cheating last year by opening their papers too early, and that others did coach their children intensively for the tests. Furthermore, children who were found to be four or five years apart in reading age achieved almost identical scores in the English comprehension test.
Dr Marks should also have recognised that a large part of the reason for the huge discrepancy in the results is that schools are not given the same "raw material" to work with. He accepts that "some higher achieving schools may well have a high proportion of pupils of above-average ability". But this understates the problem. During the past year Durham University's PIPS project researchers have collected statistics on the number of letters that children know when they enter school because this is a crucial predictor of future performance. Their 264-school survey showed that in 54 per cent of primaries newly-enrolled pupils knew fewer than five letters on average whereas in the top 16 per cent of schools children knew four times as many letters.
Baseline statistical evidence of this kind must be treated with at least the seriousness of the key stage 2 data. The Government is right to say that targets must be set for every school, but the current focus on end-of-school results and the publication next March of the 1996 KS2 test results should not divert attention from Reception and Year 1. The "Success for All" programme directed by the highly-regarded Baltimore academic Bob Slavin has demonstrated that if there is "zero tolerance" of reading failure in Year 1, and extra coaching available for those who are in danger of falling behind, a school and its pupils can blossom in the most inhospitable environments. It is an idea worth trying.