Uta Frith sighs heavily. She is talking about the decades-long debate over whether children with autism should be educated in mainstream schools. Her research has led her to reach a somewhat uncomfortable conclusion.

The emeritus professor of cognitive development at University College London recognises that the debate is polarised and acknowledges that some children do very well in mainstream education, either when they are young or when they are older. Despite this, she believes most mainstream schools are simply not set up to provide the level of personalised provision from which children with autism can benefit.

“If you want to pin me down, and I will make a lot of enemies, I would say that special education, in the long run, does seem to be the right thing,” says Frith. “You need teachers who know what autism is, know about the condition, really care about it and can deal with the different individualities in autism.

“Everyone says there is a continuum from very mild to very severe. But it’s not really a continuum - it’s like every individual is completely different. So, you need to adjust what you do for every child, every person.”

Until mainstream schools have access to the funding and expertise needed to adopt the kind of “tailor-made” approach to teaching each and every child with autism that Frith is talking about, she will be coming down on the side of special schools.

Opinion swings

Frith does not reach this conclusion lightly; her thinking has been informed by 50 years of research into autism-spectrum disorders.

During the decades she has been working in the field, Frith has witnessed “absolute swings” of opinion, not only about where children with autism should be educated, but over whether they should be educated at all.

In 1968, when Frith was completing her PhD on autism, the condition was still thought to be incredibly rare and something that appeared exclusively in children who had severe learning disabilities. It was only gradually realised that the diagnosis could apply to children with many different levels of ability.

This was also a world before the 1970 Education Act deemed that all children were educable regardless of disability. At the time, children with an autism diagnosis were generally not afforded the right to an education. For Frith, the act was a “tremendous triumph” and something that has informed the way she thinks about educating children with autism to this day.

“I have often said, and I hold to this, that education is the only therapy we have for autism. And it works,” says Frith. “But education would be defined in a very broad way. It would start at preschool age and it would not end at 18; it would go on throughout life. It would be work in progress continuously.”

Education can work for children with autism, Frith is certain of that. What is harder for her to pin down is exactly what effective autism education looks like − largely because of those extreme individualities that she sees within the condition.

“I have seen a lot of good practice,” says Frith. “I have been really deeply admiring of what teachers do almost by instinct. In a sense, this makes it very difficult to tell others what to do.”

Another complication when it comes to pedagogy, she says, is that the same approach can have very different outcomes when used by different teachers.

If, for example, you were to observe an excellent teacher of students with autism in a bid to improve your own practice, you would most likely take note of the things they did and then attempt to replicate this in your own classroom.

However, Frith says it would be all too easy to imitate the wrong aspects of what you have observed, and that even if you exactly copy another teacher’s methods, or precisely recreate something you have read about in a book, you cannot necessarily assume that the same methods would work just as well in your hands.

“It’s not only individualised learning; we need individualised teaching as well,” says Frith.

This belief in individualised teaching reinforces Frith’s thinking that special education may be the way forward. It also makes her wary of any method that claims to be the “right” approach to teaching children with autism and means she would always advise parents not to fight too hard for, or spend too much money on, ensuring that their child has access to the latest, most widely celebrated programme.

“In all these packages and programmes for autistic children, there’s an enormous amount of hype, there’s a great amount of snake oil. It really has done a lot of damage, I think, because parents have become incredibly poor from trying supposedly highly promising methods.”

The difficulty is that while a promising method might work well for one child, the same programme might have no impact, or even a negative effect, on another − and it is not easy to predict which way it will go.

Training is a red herring

At the moment, she says, most teachers do have a good enough understanding about the idiosyncrasies of autism to be able to navigate through the hype and advise parents about which methods might be best suited to a particular child.

This seems to suggest that more training for teachers could be the answer to improving provision, but Frith doesn’t think it’s that simple. In her eyes, demand for extra training is simply another red herring in the search for a solution.

“Everybody says that they didn’t get enough training. But what is enough? I think it’s an easy thing to complain about,” she says.

According to Frith, the answer to a more individualised approach to autism education - and thus better education for children with autism - is not just more training for teachers or rolling out the newest approach being hyped in the media; it’s science.

“Rather than complaining about lack of training, I complain about not enough science,” says Frith.

“There’s not enough research into what makes a good teacher and how you can match up method and teacher; teacher and child; method and child. It’s very subtle and difficult, but it should be done.”

Helen Amass is editorial content manager at Tes and a former teacher. She tweets @Helen_Amass