The DfE’s New Autism Evidence Review Gives Girls One Sentence in 93 Pages

On 30 July, the Department for Education quietly added a new document to its evidence library: a review of behavioural and social communication interventions for autistic children and children with social, emotional and mental health needs in mainstream schools. If your daughter has ever been put through a social skills programme, a behaviour chart, or an intervention designed to make her look less autistic, this document is about her. Except it isn’t. In 93 pages about autism interventions, girls get exactly one sentence, and the sentence is an admission that the evidence barely includes them.

We are quick to say when the system gets it wrong, so let’s be equally quick when government-commissioned research gets something right. This review says things we have been saying for years, and it says them in the DfE’s own publication series.

It names masking as a harm: interventions “that focus narrowly on observable behaviour or social communication may encourage masking or suppression, particularly among autistic pupils,” and “masking can also lead to poor mental health outcomes for autistic individuals.”

It asks whose outcomes actually count: “It is important to distinguish between outcomes that are genuinely pupil-led and those that are imposed by adults, where reducing a behaviour or improving social skills reflect adult priorities rather than the child’s own goals or wellbeing.” Read that again. A DfE-published review is saying that a child behaving more conveniently is not the same thing as a child doing well.

And it frames behaviour the way we do: difficulties arise “not simply because a child lacks a skill, but because there is a mismatch between the child’s profile and the demands, expectations, or affordances of their environment.” Behaviours labelled disruptive or withdrawn “are often responses to perceived demands, uncertainty, sensory overload, or stress.”

That is the environment being named as the problem, in black and white, on gov.uk. Parents fighting schools over “behaviour”: bookmark it.

Much of the research the review covers is rooted in applied behaviour analysis, because that is what dominates the published literature. But the review does not endorse it. It notes that these approaches “are controversial among educators and autistic individuals,” and that improved outcomes “may result from masking because of negative reinforcement, which is also linked to mental health issues later in development.” It cites research on autistic people’s own experiences of ABA, and its recommendations call for neurodiversity-affirming approaches that avoid promoting masking.

Positive behaviour support gets exactly one passing mention in 93 pages: a single example inside a review of social skills interventions, “individualised positive behaviour support informed by functional behaviour assessment”, with positive effects on social development reported (p. 49). There is no evaluation of PBS as a framework, and no engagement with the concerns autistic people raise about it. One line, uncritical, is worth keeping an eye on – but it is a line, not an endorsement.

None of this stops a time-pressed SENCO skimming the summary tables and concluding that an intensive behavioural programme “works.” The risk, as ever, is in how the document gets used.

Here is everything the review has to say about girls, in full: “most participants were male, and evidence is limited for female pupils” (p. 62). That is it. One sentence, in one intervention section, in 93 pages. The word “girls” never appears; “gender” never appears; the DfE’s own reviewers concede in a single line that the evidence base barely includes girls, and then the review moves on as if that were a footnote rather than the hole beneath the whole document.

No internalised presentations. No selective mutism. No emotionally based school non-attendance. No demand avoidance. No burnout. The entire review is about children who are in mainstream classrooms displaying behaviour that adults can see and measure. The girl who sits quietly, holds it together until 3.30pm and falls apart in the car does not appear in this evidence base, because the evidence base was never built to see her. Interventions are judged on observable behaviour, and her whole survival strategy is making sure there is nothing to observe.

Here is what makes it stranger: the DfE already knows. The autism identification review published in the same series last September acknowledges that autistic girls may be under-recognised, that diagnostic criteria are skewed towards male presentations, and that masking is more prevalent in girls. The knowledge is sitting one document over. It just didn’t make it into the review about what schools should actually do.

The review’s scope is mainstream settings. That is a reasonable boundary for a research project, but it means the tens of thousands of young people who are no longer in any classroom – the ones this system’s failures produced – are outside its frame entirely. There is no evidence here about what re-engagement looks like for a young person who hasn’t crossed a school threshold in a year. The children most failed by the standard playbook are, once again, not in the study.

If you are a parent: pages 13 and 14 of this review are now your citation. When a school proposes an intervention aimed at making your child’s behaviour more convenient, you can ask, in the DfE’s own words, whether the outcome is genuinely pupil-led or imposed by adults, and what the masking cost will be. That question now has government-published evidence behind it.

If you are a SENCO, an EP or a commissioner: read past the intervention tables to the discussion. Then ask why your setting’s evidence for “what works” contains no girls – and what that means for the ones on your caseload who seem fine. You may want to signpost them to our school training.

The girls missing from this review are the ones Autistic Girls Network exists for. Our training for schools and services covers exactly what the evidence base leaves out: internalised presentations, masking, and what support looks like when there is nothing to observe. And if you are the parent reading this late at night, our community of 36,000 families is where the missing evidence lives — in lived experience. Come and find us.