Many heads would give almost anything to ensure that 81 per cent of their pupils pass the all-important threshold of five good GCSEs, including English and maths. But imagine if a third of those same children scoring top results had special educational needs. Surely this huge success, the holy grail of education, is only a flight of fancy?

And yet it is not. A TES investigation has found that that is exactly what pupils achieve at several secondaries around the country. Many are beginning to ask how children with serious learning difficulties manage to defy the odds to do well in tough public exams. But the bigger question, observers are gently muttering, is whether these children really deserve to be on the SEN register at all.

Increasingly loud whispers from experts, governors and teachers suggest some heads might be “playing the system” and claiming to have more SEN pupils in order to climb up the league tables that aim to measure the context of the pupils that schools are working with - the raw materials of secondaries. If a school has a high proportion of SEN pupils together with high academic results, this can play a big part in getting good outcomes in the controversial “contextual value added” (CVA) measure.

This system of assessment was launched four years ago as a way of helping schools in poorer areas, which struggled to articulate their success purely through GCSE results. Those with more disadvantaged pupils, or extra children with special needs, are able to rank highly on the league tables if children make good progress.

But there is a problem. Since CVA’s introduction, the percentage of pupils registered as having SEN has increased dramatically. These children don’t have statements, the majority having been assessed by teachers. And it seems that “middle-class” schools and even grammars are making use of CVA too. There are now 46 secondary schools around England which claim that more than half of pupils have SEN.

In a fog of uncertainty about the rules and management, there is widespread agreement that the calculation of CVA has become a dark art, with the results labelled as “misleading” and even “dishonest”. Even the Government admits there are “perverse incentives” for schools to label more children as having SEN.

Just look at the figures. Back in 2006, the year CVA was introduced, 9.1 per cent of children in the UK were classed as having SEN. A further 8 per cent had statements or were on “school action plus” - where they get help from outside experts in school. By 2009 this had climbed to 13.5 per cent and 10.8 per cent respectively.

In 2006 there were 1.29 million children with special needs on school action or school action plus. By 2009 this had grown to 1.43 million, despite falling school rolls.

“There are probably schools which put children in the SEN register to support their CVA score, and to support their budget,” says Lorraine Petersen, chief executive of the National Association of Special Educational Needs (NASEN).

One of the big problems, according to NASEN, is that there is no national process for identifying children with SEN.

“Every school has their own individual way of doing things, but if high-performing schools have high levels of SEN, local authorities and school improvement partners should be asking questions,” Mrs Petersen adds. “Pupils with severe learning difficulties are not going to achieve at a high level in their GCSEs.”

Schools were first assessed using a “value added” measure in 2002. But four years later CVA was introduced because ministers wanted something more sophisticated. It takes into account SEN, first language, ethnicity, measures of deprivation and of pupil mobility, the number of pupils in care and the spread of abilities within the school.

Concerns about the validity of CVA started appearing just two years after it was introduced. Bristol University researchers said the information was meaningless, with some of the data “at best misleading, at worst dishonest”.

Up and down the country there are examples of where SEN numbers are counter-intuitive. Poole High School in Dorset, for example, had 6 per cent of pupils on the school action register last summer. But the local selective school, Poole Grammar, had 11.2 per cent - and a higher CVA grade. In 2009, 99 per cent of 16-year-olds at the grammar school got five good GCSEs, including English and maths.

The school has now won specialist SEN status, awarded after much persuasion by head Ian Carter. “All kinds of children can have extra needs, even if they are gifted and talented,” he says.

Pupils on the school’s SEN register are dyslexic, autistic, deaf or have ADHD.

“We want to do the best for our pupils, we’ve found this is simply good teaching. I think our intake shows grammar schools are not just middle class enclaves,” Mr Carter adds. “I’m not concerned about our CVA grade, we don’t attach much value to it.”

The Bristol study - by Dr Deborah Wilson and Anete Piebalga - found that because almost half of English secondaries’ CVA results are indistinguishable from the national average, a very small difference can leave schools hundreds of places apart in the tables.

These concerns are clearly shared by the Department for Children, Schools and Families. This year, senior SEN advisor Philippa Stobbs admitted there were “perverse incentives” for teachers to label pupils as having special needs and this then hinders their achievement.

Ms Stobbs said “over-labelling” was an example of “laziness” in mainstream schools. She called for Ofsted and school improvement partners to question primaries and secondaries with very high numbers of SEN pupils.

A recent report produced for the DCSF by researchers at the National Children’s Bureau, the Thomas Coram Research Unit, the Institute of Education and the Council for Disabled Children, also found there were “perverse incentives” for teachers to identify children with SEN. These included extra funding as well as a better CVA measure.

“Schools are very honest and say that the more children they’ve got on their SEN register, the better their value added looks, so there are a number of things around that we’ve got to overcome if we’re going to get an honest reporting in terms of SEN in schools, I think,” a senior SEN and inclusion advisor told the researchers.

Indeed, a county council SEN and school improvement manager told the researchers: “There is a slight incentive for schools to identify children in terms of accessing funding that is then delegated, because higher incidence of special needs will bring more funding ... Some schools, I think, have kept children at school action when they haven’t needed to because they’ve recognised it has an impact on their achievement results.”

Another council assistant director of inclusion accused teachers of “over identification”: “I think in some of our schools they’ve lost the plot as to what the normal ... is, and that’s driven by (attainment) targets, it absolutely is ... Where a school is struggling, and some of our schools are struggling, it’s quite a neat way of saying, it’s not about the teaching and learning, it’s to do with the child.”

The report found too many children who were just underachieving were being labelled as having SEN. Other heads faced with an influx of children who they knew wouldn’t achieve were over-identifying in an attempt to not jeopardise the school’s attainment results.

Now DCSF officials seem to want to close any loopholes. Experts at Oxford University are trying to come up with a way of working out a more foolproof version of CVA, less likely to contain errors.

Christopher Robertson, lecturer in inclusive and special education at Birmingham University, says heads are not investigating the causes of over-identification of special needs.

“Many will acknowledge their SEN levels are high, but they don’t always question why that’s the case,” he says.

“Most just blame their catchment area, but don’t want to probe too much.”

The Government first started demanding detailed data about SEN from schools in 2005. Mr Robertson says this has also made schools more likely to “categorise” children.

“In many cases this labelling has been a retrograde step,” he adds.

Harvey Goldstein, professor of social statistics at Bristol, says the Government must accept that a measure like the CVA will always be manipulated.

“It’s taken the political stance that you need league tables, but the rankings will always be imprecise,” he claims.

“What’s happening with SEN is a side effect, in a way schools are the victim so it’s no wonder they choose to play the system. They will always do this if it’s in their interest to do so and if the national policy encourages them to do so.”

Simon Uttley, head of St Paul’s Catholic College in Sunbury, Surrey, says school leaders are under “increasing pressure” to improve results.

“The feeling is the new Ofsted framework will mean more importance attached to the CVA, a low score could involve going into a category,” he explains.

“But heads feel if they don’t identify SEN they are doing children a disservice, they are just doing their job.”

Whatever is responsible for the boom in SEN numbers - conscientious teachers or a desire to play the CVA system - it seems ironic that a measure designed to introduce much-needed fairness could be the victim of manipulation. Whether or not children benefit from CVA is another story altogether.

‘PROGRESS IS EXCEPTIONAL’

Four years ago 1.9 per cent of pupils at Dover Grammar School for Girls were classed as having special educational needs, and 1 per cent had statements or were receiving other professional help.

By 2009 14.7 per were on the SEN register. The school has a contextual value added score of 1,030, which puts it in the top 20 per cent nationally. Despite this, every single GCSE pupil got five good grades including English and maths.

These results are down to hard work by staff and students, head Judith Carlisle says, and SEN numbers have not been inflated.

“We serve a coastal town, an area of high deprivation. Any assumption that we wouldn’t have children with special needs is ridiculous. We are Dover, not Tunbridge Wells,” she says.

“The CVA is important. Ours is high because children arrive having scored lower at key stage 2. The progress they make here is exceptional.”

The school’s ethos is to identify SEN early, then to teach children in small groups. Teachers work closely with any child who has problems learning.

THE COST OF INCLUSION

In 2009 almost 43 per cent of GCSE pupils at St Marylebone CofE School in London had special educational needs - but 81 per cent got five top results, including English and maths.

The school has a value added score of 1,035. But head Elizabeth Phillips says this is irrelevant.

“Parents are not bothered about it, what they are interested in is the main outcome, GCSE results,” she says.

“I don’t put a lot of value in CVA, I’m not convinced Ofsted does either.”

Identification of learning problems is a priority for teachers at St Marylebone, which is a specialist SEN school. Those who need help get taught by a teacher, rather than a teaching assistant, in smaller classes. Mrs Phillips says this is the reason for the school’s incredible progress, but that operating in this way costs money.

“The fact children have statements helps to fund this, but it doesn’t pay for it all, we have to put in extra funding,” she said. “Tony Blair started the move towards inclusion, but I don’t think he took into account the cost of this. It’s expensive, and unless you fund it properly SEN pupils get an inferior education at a mainstream school.”