School experiences and inclusion

International policy and a global trend favor teaching autistic pupils in mainstream schools.¹ Within a mainstream school, some autistic pupils spend part of the week in a specialist resource center, a base staffed by people trained in autism, while others are taught only in mainstream classes; some autistic pupils attend a separate special school.²

Whether mainstream placement helps has no simple answer. A Campbell review pooled 15 studies of children with special educational needs and found that inclusion slightly favored children's learning and adjustment, but that no result was statistically significant; the authors concluded it is very unlikely that inclusion in general improves or harms learning and psychosocial adjustment.³ In a French cohort of 356 autistic children followed for three years, children taught in special-education classrooms had lower communication and daily-living skills at the end than those in regular classrooms, with no difference in socialization skills or IQ. The authors reported wide variation between individual children.⁴

A three-year study in England followed 119 autistic pupils aged 11 to 14 across three placements: a specialist resource center, mainstream classes in a school that had a resource center, and mainstream classes in a school without one. Placement type on its own made no difference to the pupils' psychosocial outcomes. Pupils in a resource center had higher attainment and a stronger sense of belonging than those in mainstream schools with no center, and felt more supported by their teachers.² Across all the models, the strongest predictors of doing well were feeling supported by teachers and by other pupils.² A separate case study of seven secondary schools described resource centers as predictable, trusted places, but found that staff shortages and thin mental-health services limited even the schools that were doing well.⁵

Studies of single points in time point in different directions. In one, of 860 autistic children aged 3 to 14 receiving rehabilitation services in China, the 36% who attended mainstream school reported higher social inclusion, most clearly those with milder traits.⁶ In another, 137 autistic adolescents without intellectual disability in special-education classes reported more social anxiety, avoidance and loneliness than those in general classes; class type also shaped how closely those feelings hung together, and the authors noted that their design cannot show whether the placement caused the differences.⁷ One meta-analysis found that autistic students in inclusive mainstream settings were victimized more, not less, than those taught separately,⁸ so placement cuts both ways. A review of 233 studies found that research on inclusion more often asked the autistic pupil to adapt to the school than it adapted the school to the pupil.⁹

On average, autistic children and adolescents take part less in peer groups, sit more on the edges of them, and are accepted less by other pupils than non-autistic children.¹⁰ That does not mean they want less contact. In one study of 47 autistic and 52 neurodiverse non-autistic pupils aged 8 to 13 in two special-education primary schools, the non-autistic children felt lonelier the less time they spent with others, while the autistic children felt lonelier when other pupils did not want to play with them.¹¹ In a study of 241 autistic young people aged 15 to 26, loneliness went with feeling bad while interacting and with being ignored by peers, not with how much time was spent alone or with others.¹²

In a study of 306 young people aged 9 to 16, of whom 86 autistic boys and 18 autistic girls, the autistic young people reported fewer positive features in their friendships than non-autistic peers, such as closeness and support.¹³ For the autistic girls alone, more positive friendship features went with more anxiety rather than less. Even good friendships can carry a cost: in a cross-sectional survey of 72 autistic students in mainstream secondary schools, camouflaging accounted for the statistical link between a sense of belonging and anxiety.¹⁴ For 224 autistic young people aged 15 to 25, being satisfied with the support a friendship gave softened the effect of being victimized on depression.¹⁵

What pupils want from friendship varies. In interviews with 19 autistic children in UK primary schools, many valued companionship and play more than the close, confiding friendship adults often expect, and not all wanted friendships of the kind their non-autistic classmates had.¹⁶ Autistic young adults who had left mainstream education were positive about support from other autistic pupils.¹⁷

Autistic pupils are bullied more than non-autistic pupils. A meta-analysis of 17 studies put the pooled rate of victimization at 44%, with 10% involved in bullying others and 16% in both roles.¹⁸ A second meta-analysis of 34 studies put the pooled rate of victimization at 44% and of bullying at 47%; it pooled autistic children and adults together and counted all kinds of victimization, not only school bullying.¹⁹ A third review of 34 studies put victimization at 67%, higher than both non-autistic students and students with other disabilities, and found that autistic students in inclusive mainstream settings were victimized more, not less.⁸ The three estimates differ partly because they pool different definitions, ages and settings, so the range runs from 44% to 67%.

Large population studies agree on the pattern. In a UK cohort of 576 autistic and 14,963 non-autistic young people followed from age 5 to 14, 74% of the autistic group fell into a bullying trajectory at some point, against 38% of the non-autistic group; being involved in bullying was associated with poorer mental health at 17 in both groups.²⁰ In a Finnish sample of 4,408 eight-year-olds, 46% of the autistic children were victimized, against 2% of the whole sample, and victimization rose with autism trait scores.²¹ In a US national survey of 71,973 young people aged 12 to 17, the rise in anxiety and depression linked to being bullied was greater for autistic young people than for others.²²

Autistic-led research describes barriers that come from how schools are run, not only from individual staff. In a community-participatory study, a mostly autistic research team interviewed 19 autistic adolescents and adults about self-advocacy at school. The participants described six kinds of structural barrier: erasure, conformity, isolation, oppression, hidden expectations, and authority.²⁴ They described these as features of the system rather than the failings of particular teachers.

The pressure to conform shows up in how pupils are expected to behave. In 31 autistic girls in general-education high school classes, camouflaging was most frequent in those general-education classrooms, and with teachers and classmates they did not know well; girls diagnosed later camouflaged more and reported more costs.²⁵ In a survey of 248 autistic young people, about two-thirds said autism had a negative effect on their high school experience, while the majority rated the impacts of school professionals, family and peers as positive.²⁶

The research on inclusion has often reproduced this gap. A review of 24 studies in which autistic pupils helped shape school research found that few reported feeding their findings back to the schools, so it is hard to tell how far any of it reached classrooms.²⁷

Noisy corridors, canteens, strip lighting and crowding are part of the school day for everyone, but for many autistic pupils they are disabling. In a participatory study in which six autistic students aged 12 to 16 took photographs of their school, the students described being routinely overstimulated and short of support, and stressed how much they needed quiet places to recover; the authors recommended quiet rooms and cheap changes such as replacing fluorescent lighting.²⁸

In a study of 26 autistic and 26 non-autistic young people aged 8 to 14, sensory processing was associated with school performance, and school competence was lowest among the autistic children with the greatest sensitivity to sensory input and the fewest ways of avoiding it.²⁹ A systematic review of accommodations in mainstream schools for children on the autism spectrum found the evidence inconclusive: of 37 studies, only 14 met basic quality standards, and those tentatively suggested that changes to the teaching and social environment can help.³⁰ The largest review of school-based inclusion interventions found that very few of the 56 studies tested a change to the physical environment at all.¹

Interventions that aim to include autistic pupils usually change the pupil rather than the school. A systematic review of 56 school-based interventions found that they made school activities more accessible, but did not reliably improve the give-and-take of friendship between autistic pupils and their classmates.¹ Social skills training is weaker than it looks: a meta-analysis of 49 randomized trials found that once the results were adjusted for the likelihood that positive findings were published and negative ones were not, the programs had no overall effect, though small effects appeared when the results were split by who reported them.³¹

Some approaches have more support, though it is early. A review of peer-mediated programs in primary school found they improved social skills in the short term, but only 5 of 24 studies checked whether the gains lasted, and those that did disagreed.³² A review of peer-support programs for autistic people aged 12 and over reported better well-being, self-esteem and schoolwork, but the programs had small samples and no control groups.³³ Training staff matters, though the clearest test so far used university educators: a co-produced autism and universal-design training improved their understanding of autism, and they still remembered it a month later.³⁴

Autistic adolescents describe what helps them day to day: a teacher they can trust, being given choices through accommodations, and work built around their interests.³⁵

Teachers often start without much preparation for autistic pupils. In a survey of 369 primary teachers in Ireland, most had little autism content in their initial training and almost no further training before teaching an autistic child.³⁶ In a survey of 887 general-education teachers in Germany, knowledge of autism was moderate, confidence was not high, and attitudes toward inclusion were fairly positive.³⁷ A systematic review of 57 studies found that teachers' confidence was held back by limited professional development and by structural barriers such as time and support.³⁸

One source of missed support is how autism is recognized. In an experiment, primary school teachers read descriptions that varied both the child's gender and whether the description matched a male or a female picture of autism. The teachers were less likely to identify autism in girls and rated the boy as more likely autistic, and the effect of gender appeared only for the female-style description.³⁹ Training is linked to confidence: in six secondary schools, autism-specific training and neurodiversity-affirming attitudes went with greater staff confidence.⁴⁰

Autistic writers describe school as a place built for someone else. Pete Wharmby writes that schools "have been designed with aims and ambitions in mind that are not well suited to the neurodivergent minority", and that narrow corridors, strip lighting and a lack of quiet places make the building itself a barrier.^c1 Carly Jones, an autistic writer on safeguarding girls, writes that sensory difficulties are "an accessibility need" rather than a preference or a fuss, and compares a school that ignores them to one without a ramp.^c2 Wharmby argues that schools spend their energy expecting autistic children "to somehow adapt to intolerable environments" and never consider changing the environment instead.^c1

Katie Levin describes how an ordinary school routine can leave an autistic pupil on the outside. When a teacher asks the class to pick a partner, she writes, "Since I was too afraid to ask anyone (and no one ever asked me), I was almost always one of the leftovers."^c3 Being quiet and getting good grades, she adds, does not mean school is going well.^c3 Carly Jones adds a practical point: a unit for autistic pupils inside a mainstream school is a model she believes all mainstream schools should offer.^c2

Several writers in Leaders Around Me, who type or spell to communicate, describe wanting the ordinary parts of school life. Coby Khodosh, then nine, writes that he "craved more, I wanted to belong".^c4 Emmanuel Bernabe, twelve, writes of starting school "scared that I would be lonely, unsure that people would try to understand me", and of the relief of being "surrounded by friends".^c5