For autistic pupils, the school matters less than the relationships inside it

The debate over how to educate autistic children has been framed, for decades, as a question of placement. Mainstream or specialist? Inclusion or separation? A new three-year study from the University of Surrey suggests this framing may be missing the point.

Placement versus experience

Researchers followed 119 autistic pupils aged 11 to 14 across seven mainstream secondary schools in the United Kingdom. The students were divided into three groups: those attending a specialist resource centre (SRC) within a mainstream school, those in mainstream classes at a school that hosted an SRC, and those in mainstream schools without any specialist provision. The study compared outcomes across psychological wellbeing, social inclusion, school belonging, academic progress, attendance, and exclusion.

The specialist resource centres did show certain advantages. Students placed in SRCs had higher academic attainment, a stronger sense of school belonging, higher perceived teacher support, and smaller attendance gaps relative to school-wide averages compared to national benchmarks for autistic pupils.

But here is the finding that complicates the placement debate: autistic pupils in SRCs had similar overall levels of psychological wellbeing to those in other types of schools. The type of provision did not determine how well students were doing emotionally.

What actually predicted outcomes

Across all three school settings, the strongest and most consistent predictors of positive adjustment were not structural - they were relational. Perceived peer support was most strongly associated with reduced internalizing symptoms, fewer peer problems, greater happiness, stronger friendships, and increased flourishing. Perceived teacher support was the strongest predictor of school belonging and was also linked to increased flourishing and fewer peer difficulties.

In other words, whether a student felt supported by the people around them mattered more than where they were placed. A student in a mainstream school who perceived strong peer and teacher support could thrive. A student in a specialist centre without that support could struggle.

Why SRCs may still help

This does not mean specialist resource centres are irrelevant. The study's lead author, Dr. Anna Cook, a developmental psychologist at the University of Surrey, noted that mainstream school environments can be overwhelming for some autistic pupils due to noise, sensory demands, and social pressure. SRCs offer quieter spaces with higher staff-to-student ratios, which may help some students stay engaged in education.

The SRC advantage in academic attainment and belonging is real and meaningful. These are not trivial outcomes. For a student who might otherwise disengage entirely from school, a specialist centre could be the difference between continuing education and dropping out.

But the data suggest that the SRC's benefits may operate largely through the relationships it facilitates. Higher staff ratios mean more teacher attention. Smaller groups mean more manageable social environments. The structural features of the SRC may matter primarily because they create conditions where supportive relationships are more likely to form.

Policy implications and limits

The implications for education policy are direct. Decisions about where to educate autistic students are frequently made at the system level - based on available placements, budgets, and bureaucratic categories. The study suggests that these placement decisions, while important, are incomplete. Without investment in the quality of relationships within whatever setting a student occupies, structural reforms alone will fall short.

The study has notable limitations. The sample of 119 students across seven schools is relatively small. The research was conducted in the UK, where educational structures and support systems differ from those in other countries. And the study measured perceived support - what students reported feeling - rather than objectively quantified teacher or peer behavior. Perception and reality may diverge.

The three-year timeframe is a strength, providing longitudinal data rather than a snapshot. But it covers only early secondary school years (ages 11-14), and outcomes may shift as students face the increasing academic and social pressures of later adolescence.

Still, the central finding is hard to dismiss. Improving outcomes for autistic students may depend less on which door they walk through each morning and more on what they find on the other side of it.