Firstly, Professor Gorard himself makes it clear that there are many factors in determining the nature of a school’s population and whether these factors reflect the socio-economic nature of its neighbourhood. We agree with this. The most important of these factors is whether a school has autonomy over its own admissions policy. However, the study focuses heavily on foundation and voluntary-aided specialist schools which have this autonomy but now account for only a third of specialist schools.

The first cohort of specialist schools in 1994 were all grant-maintained and voluntary-aided since only those schools could apply. This policy was changed in 1995 and in recent years the proportion of community schools has risen dramatically and they now make up over 65 per cent of the overall total of 700 specialist schools. Thus, nearly two-thirds of specialist schools’ admissions are determined by their education authority.

Moreover, the “segregation ratio”, on which Professor Gorard bases his measure of “privilege”, is based on the proportion of children eligible for free school meals. This is a notoriously unreliable measure as many parents, especially Asians, prefer not to claim these. Many other parents are in low-paid jobs which make them only just ineligible. But even on the measure of free school meal (FSM) eligibility, the current 700 specialists now have a FSM average of 17.4 per cent. This is slightly more than the national average of 16.5 per cent for pupils in all maintained secondaries, according to OFSTED figures for 2000.

Nor is academic selection the key to specialists’ success. Professor Jesson’s study of their outcomes last year shows that in both 1996 and 1997, when reliable key stage 2 data was available for all schools, there was only a small difference in the average KS2 point count of Year 7 pupils arriving at specialist schools, compared to other comprehensives (25.4 per cent for specialists versus 25.1 per cent for others in 1996; and 25.9 per cent versus 25.4 per cent in 1997). Better results at GCSE for specialist comprehensives in 2000 (53 per cent average five-plus A* to C at GCSE, compared to 43 per cent for other comprehensives) is not, therefore, dependent upon selecting more able cohorts.

Schools’ performance should also be judged on a comparison of their ability intake at age 11 as measured by KS2 results with actual GCSE results five years later.

Professor Jesson’s study of specialist school results in 2000 compared predicted results at GCSE using KS2 data in 1995 (in schools that have reliable KS2 results - about two-thirds) with actual GCSE results in 2000 . He found that students in specialist schools scored 5.4 percentage points better than predicted on average. Students from other comprehensives, by comparison, scored an average of minus 1.1.

What is important is that specialist schools help to raise standards in all areas and add more value, whatever the ability of pupils. Critics say this is because of extra funding. In fact, the additional recurrent funding is relatively modest at pound;123 per capita per year. We would argue that it is the process of bidding and of keeping specialist status - the planning, the targets for each pupil, the accountability, together with outstanding headteachers, which has led to improvement above the national norm.

Sir Cyril Taylor is chairman of the Technology Schools Trust