Neurodiversity and GCSE Assessment: Rethinking Fairness in Exam Design

Assessment lies at the heart of the GCSE qualification. Yet, for many neurodivergent students, exam outcomes don’t reflect their knowledge or potential. New research funded by the Nuffield Foundation is helping us understand why that gap exists — and what it might mean for classrooms and revision practice across the UK.

The Challenge: Persistent Attainment Gaps

National data consistently shows that neurodivergent students — including autistic learners — achieve lower average GCSE results than their neurotypical peers, even when access arrangements like extra time are in place. Critics have long questioned whether the structure of assessment might contribute to these outcomes. Nuffield Foundation

The new project takes this question directly to the exam papers themselves. Rather than assuming that access arrangements are sufficient, researchers are analysing thousands of scripts and a range of question features — such as multipart structures, embedded idioms, and format complexity — to see how these factors correlate with performance differences across neurotypes. Nuffield Foundation

Evidence‑Led, Collaborative Inquiry

What makes this work distinctive is its design: it combines traditional statistical methods with novel AI analysis, and crucially, involves neurodivergent learners in co‑designing and testing alternative paper formats. This kind of participatory, evidence‑based approach helps ensure that insights translate into practical assessment improvements — not just academic critique. Nuffield Foundation

The early stages involve identifying which elements of exam design might inadvertently disadvantage some students and exploring how changes could help level the playing field. For example, simplifying question language, reducing unnecessary cognitive load in multipart items, or offering alternative ways to demonstrate understanding are all areas under consideration.

Practical Implications for Schools and Revision

For teachers and parents, these findings should influence how we prepare neurodivergent learners for high‑stakes assessments:

1. Teach how to read questions, not just content. Explicit practice in parsing complex prompts can reduce unnecessary misunderstandings.

2. Diversify practice formats. Using a mix of short‑answer, structured response and project‑like tasks helps learners build confidence across formats.

3. Develop metacognitive strategies. Encouraging students to reflect on how they approach questions — planning, checking and pacing — can reduce anxiety and improve performance.

Preparation that mirrors exam conditions — in how questions are framed — builds both familiarity and resilience.

Assessment Fairness Is Not Static

The broader message is clear: fairness in assessment is not an afterthought; it is a design problem. If evidence suggests that some question features systematically disadvantage groups of learners, then educators, exam boards and policymakers have a responsibility to examine and redesign those features. Nuffield Foundation

Neurodiversity doesn’t mean deficit; it means diversity. Assessment systems that recognise and respond to that diversity — grounded in evidence and co‑designed with learners — will serve all students better.