During the noughties, the medic and science writer Ben Goldacre became well-known for demystifying epidemiology and challenging antivaxxers long before it was either fashionable or necessary.

But it was his frequent and often funny takedowns of the “pseudoscience” of Brain Gym - an “educational movement-based model”, popular in many primary schools at the time - that first brought him to the attention of teachers.

Arguing for the need for teachers to be better able to spot peddlers of pedagogical snake oil, in 2013, he called on schools to stage the kind of evidence-based practice “revolution” that had transformed his profession of medicine.

“There is a huge prize waiting to be claimed by teachers,” he began in a paper for the Department for Education, titled Building Evidence into Education.

“By collecting better evidence about what works best and establishing a culture where this evidence is used as a matter of routine, we can improve outcomes for children and increase professional independence.”

Central to his pitch was a proposal to greatly increase the number of randomised controlled trials (RCTs) conducted in schools. Widely regarded as the “gold standard” for evaluating the effectiveness of interventions, RCTs were, Goldacre argued, the best way of identifying “what works best”.

Despite initial pushback from some teachers and academics over Goldacre’s method, doing “what works” - applying evidence from gold-standard research to decision-making - has become an instrumental force in education and other areas of public policy and practice.

Before the pandemic, the deployment of “what works” approaches did much to help narrow the attainment gap between disadvantaged pupils and their peers.

The compelling evidence on the impact of one-to-one tuition was key to convincing the government to invest £1.7 billion in a National Tutoring Programme to address learning lost to lockdown for those eligible for the pupil premium, and to spend a further £17 million on a catch-up scheme for children in Reception whose development of speech, language and communication skills had been affected by the pandemic.

The “what works” movement has transformed how schools understand and use research evidence to improve outcomes, especially for disadvantaged pupils - a target constituency of the What Works Centre for Education, better known as the Education Endowment Foundation (EEF).

Important, impressive and rapid though the growth of “what works” has been, it has a predominantly mainstream outlook.

This prompts questions about whether special educational needs and disabilities (SEND) has benefited from, or been left behind by, this influential movement.

A decade on, it is time to ask: does “what works” work for SEND?

SEND in mainstream schools

The conclusions from meta-analyses, systematic reviews and other robust syntheses of quality evidence on effective teaching for pupils with SEND in mainstream settings are consistent.

David Mitchell, author of What Really Works in Special and Inclusive Education: Using Evidence-Based Teaching Strategies, says those with SEND “respond to universal approaches in much the same way as non-SEND pupils”.

Put another way, good teaching for pupils with SEND is good teaching for everybody.

“It’s the quality of everyday activities and interactions that drive development,” says Mairi Ann Cullen, a former senior research fellow at the University of Warwick, who led an evidence review commissioned by the EEF.

A 2018 study by US academics, which is cited in the review, specifies 22 “high-leverage practices”, including explicit instruction, cognitive and metacognitive strategies, and scaffolding, which have been found to be effective for pupils with “mild disabilities in general education classrooms”.

Gary Aubin, SEND leader for a multi-academy trust and EEF Associate for SEND believes these are strategies that teachers can add to their practice relatively easily.

He says “teachers should find it reassuring” that the first line of defence in meeting the learning needs of pupils with SEND is the day-to-day high-quality teaching techniques in which they are knowledgeable and skilled.

“There’s a role for bespoke, individualised approaches,” he says of the practical challenge facing mainstream teachers.

“But it’s unrealistic to expect teachers to put in place something different for every pupil with SEND.”

Jess Howard, special educational needs coordinator (Sendco) at Laureate Academy in Hertfordshire, agrees, adding that label-driven thinking does not help.

She encourages and supports teachers in her school to develop a repertoire of proven strategies that can address the everyday requirements of a broader base of learning needs and not get too distracted by a specific diagnosis.

“We get better buy-in from teachers when we say ‘this approach would benefit pupils that find communication and language overwhelming’, rather than ‘this would benefit your autistic pupils’,” she explains.

Aubin makes a similar point about the selection of interventions.

The starting point for addressing gaps in SEND provision and practice, he says, is neither a prescribed programme nor a particular practice but teachers’ knowledge of their pupils’ areas of development, and their strengths and preferences: see the need, teach to the need.

This is a useful corrective to the common perception that SEND is inherently complex and that teachers require a different inventory of strategies to teach pupils with additional needs.

“Many teachers and school leaders still look for ‘silver bullets’ that are not only specific to SEND but also specific to individual categories of SEND,” says Mitchell.

The same thinking lies behind the way in which the task of teaching pupils with SEND is often outsourced to teaching assistants, who may not have any more or better training on how to do this than teachers.

As research has shown, the consequences of this arrangement are that, compared with their classmates, pupils with SEND do less well academically and are at greater risk of developing a dependency on adult support.

Igraine Rhodes, the EEF’s head of programmes, makes the point that schools must be mindful of “the potential damage that could be done by saying that children with SEND in mainstream classes need to be treated completely differently, because we have no evidence that’s true”.

Complex needs

There is also little evidence of what works for pupils with more complex needs, who are more likely to be taught in special schools.

It is not that the picture is inconclusive but that the weight of the evidence, and the guidance produced on the back of it, talks to those who teach and support pupils in mainstream settings.

At the same time, there’s a failure to acknowledge the links between SEND in mainstream and SEND in special schools - and the grey area this creates.

“What’s out there is more for those on SEN support and less for those with an EHCP,” says Dr Nic Crossley, national SEND representative for the Association of School and College Leaders and CEO of Liberty Academy Trust, which comprises three schools for autistic children.

“The specialist sector isn’t reflected in the world of ‘what works’.”

One explanation for this is the methodological incompatibility between the “what works” brand and the methods many SEND researchers deploy in their investigations.

Another is what exactly is meant by “works”.

A methodological mismatch?

Many studies involving pupils with complex needs and specialist settings do not use the methods that are the kitemark of the gold-standard research, which forms the bedrock of “what works”.

This does not mean such studies are in any way sub-standard. It is just that they are not large-scale trials that use a randomised controlled or matched (or quasi-experimental) design, capable of producing statistically significant results.

“There are practical limitations in carrying out gold-standard research with SEND populations,” says Mitchell.

.

Individually and collectively, he explains, these populations are much smaller in size and more diverse in terms of need, compared with the larger, more homogeneous non-SEND population.

“In combination, these factors pose problems in assembling sufficient numbers to enable matched groups to be established,” he says.

As SEND researchers are essentially drawing on a limited pool, impact trials that require hundreds, if not thousands, of pupils to take part in order for the results to achieve statistical validity are rare.

That said, new research, led by Professor Jo van Herwegen at University College London, charts a steady increase in the number of studies of targeted interventions to improve educational outcomes for pupils with SEND, which use either an RCT or quasi-experimental design.

Compared with the broader acceleration in gold-standard research globally, though, the overall number of such studies is still relatively small.

For instance, the EEF commissions about one in five of the RCTs conducted in education worldwide, including some of the largest, but only one of the 200-plus trials it has completed to date has involved special schools.

The trial of Headsprout - a programme that aims to build fluency in essential early reading skills - succumbed to various pressures, including staff turnover and workload, pupil absence and time constraints, which combined to prevent it from being delivered as intended.

These factors, exacerbated by Covid, also impacted the evaluation activities.

In the end, there was a great deal of uncertainty about the conclusiveness of the results.

“The EEF gets a lot of applications to evaluate programmes aimed at the lowest performing pupils in a particular subject, helping them to access something they are not accessing,” says Rhodes.

EEF trials often absorb pupils with SEND into this wider group, she continues, but this can be problematic in the context of a rigorous gold-standard trial.

“Often you are targeting larger groups with an intervention, so you rely on a broader label,” says Van Herwegen.

“But you need narrow definitions, so that you can understand the mechanisms of what works for whom.”

What does the ‘works’ in ‘what works’ mean?

And this still leaves the question of what we even mean by “works”, in this context.

In the sphere of “what works”, “works” generally refers to the impact on academic outcomes.

This presents a problem in relation to special schools because there is no longer a standardised approach to collecting data on progress in curriculum areas for pupils in these settings.

The DfE’s decision, in 2020-21, to remove P scales - the attainment targets for pupils working below national curriculum level - makes the task of determining what works for pupils with complex SEND over time much harder, says Van Herwegen.

“We’re doing a bad job of collecting the kind of longitudinal data that we need to inform judgements on whether to use and invest in particular approaches at the school level,” she says.

The EEF prioritises academic attainment in the trials it funds, although some programme evaluations do include procedures to measure non-cognitive skills, such as self-efficacy or self-regulation.

The emphasis on academic outcomes in mainstream settings is understandable, but the specialist sector tends to take a more expansive view of impact.

As Crossley argues, so-called “softer” outcomes may play a more significant, indirect role in attainment for pupils in these settings.

For instance, teachers and support staff might have to put considerable effort into building youngsters’ reserves of self-confidence and independence in order to create the requisite platform for effective learning.

While it is not a like-for-like replacement of P scales, it is possible that the new engagement model, which is aimed at “pupils working below the standard of national curriculum assessments and not engaged in subject-specific study”, could, in time, provide a consistent means of capturing pupils’ progress towards a ready state for learning.

‘What works’ toolkits

The relative lack of gold-standard research relating to SEND means it is under-represented in perhaps the most influential “what works” resource: the EEF’s Teaching and Learning Toolkit.

The toolkit is the first port of call for many school leaders and teachers seeking effective evidence-based solutions for pupils with SEND.

“It’s a useful place to start,” says Howard.

“But you need to bring your professional judgement and understanding of your pupils and your context to your interpretations.”

For those working with pupils with complex needs and in specialist settings, the gap between what the EEF toolkit offers and what these pupils need is considerable.

This prompts practitioners to look for inspiration in the parallel “what works for SEND” toolkits produced by organisations within the SEND sector.

The Whole School SEND Toolkit, for example, collates information about interventions and strategies, including “what works” databases from SEND groups and charities, according to the four areas of need.

On the one hand, SEND toolkits allow for the use of findings from small-scale, qualitative research that are common in SEND.

On the other hand, some of these resources invite schools to submit self-reported case studies describing the implementation and impact of strategies that they claim have been successful in their setting.

The problem with this is that it distorts the meaning and value of the “what works” brand.

“When some school leaders and teachers use the term ‘evidence-based’, they are referring to something they have seen or heard works in another school,” Van Herwegen points out.

Crossley similarly worries that the drift from an authentic research-informed approach perpetuates the somewhat haphazard practice of window-shopping for potentially useful strategies - the very thing that “what works” was introduced to remediate.

The challenge ahead

So, the cautious answer to the question of whether “what works” has delivered for SEND is perhaps: not yet.

For Crossley, though, that’s not something to feel pessimistic about. “‘What works’ is not a quick fix,” she warns.

There are now good foundations on which to build.

For example, when it comes to commissioning and conducting trials relating to SEND in various settings, the EEF is aware of what it is up against - and is prepared to do more.

“We’ve learned enough from our work so far to know that we’ve got a huge challenge ahead. It’s definitely one that the EEF is up for addressing,” says Rhodes.

Crossley, meanwhile, sees potential for immediate collaboration between the specialist sector, researchers and funders.

“We could run trials of interventions that have been shown to have an impact in mainstream settings in specialist contexts, or investigate whether SEND variables affect impact,” she says.

“It might be that there isn’t anything wildly different that we need to be doing for our SEND population.

“But we do need to be taking the evidence and looking at how it applies to our context, and what adaptations are needed to make it work in different SEND settings.”

That SEND is on the radar of organisations like the EEF is welcomed by Simon Tanner, national director of SEND for E-ACT multi-academy trust.

He worries that the growing population of pupils with complex needs who require, but cannot get, a place in a specialist setting are being “fundamentally failed” by what constitutes ordinarily available provision in mainstream schools.

“There’s an urgent need to improve teachers’ pedagogy,” he says.

Mitchell echoes this, suggesting that “quality research is needed on appropriate pedagogical adjustments in the area of neurodiversity”.

In his 2013 call to arms to schools, Goldacre - somewhat mixing his metaphors - cited the progress made in his own field: “Medicine has leapt forward with evidence-based practice.

“Outcomes for patients have improved as a result, through thousands of tiny steps forward.”

That we now hear far less of the “edu-quackery” that exasperated Goldacre is one small but meaningful marker of how many more school leaders and teachers are critical consumers of evidence than a decade ago.

Goldacre brought to public attention the ways in which hopeful-sounding, usually expensive, and ultimately ineffective interventions and treatments thrive in contexts where the circumstances are particularly acute and good evidence is at a premium - but there is still more to do.

“Pupils with SEND, and their teachers and parents, deserve the best research methods to be used when identifying effective approaches to teaching,” says Cullen.

We might not be there yet, but what works for SEND is a place we can get to. And given the ongoing crisis in SEND, it is essential that we don’t give up on trying to reach it.

Rob Webster is a researcher specialising in SEND and inclusion. His book, The Inclusion Illusion, is free to download via UCL Press