The Exceptional Calculation Project
What we want to find out
Some autistic individuals show exceptional strengths in music, art or mathematics. These striking cases have sparked long-standing interest in the relationship of autism and exceptional abilities, yet the mechanisms underlying this association remain poorly understood. This study aims to:
- Characterise individual differences in arithmetic fluency and mental calculation ability across autistic and non-autistic adults using behavioural assessments.
- Identify behavioural and biological markers associated with exceptional arithmetic fluency, including measures of brain structure and function.
- Examine how autistic, systemising, and empathising traits (AQ, SQ, EQ) are related to arithmetic fluency and mental calculation ability.
- Explore whether exceptionally high arithmetic performance reflects distinctive profiles or whether it simply what is possible at the extreme end of human variation.
Why we’re doing this research
While individual case reports describe remarkable calculation skills in some autistic people, larger-scale studies show that mathematical performance varyies widely and often poorly predicted by IQ alone.
Emerging evidence suggests that factors such as verbal ability, working memory and systemising tendencies may better explain this variation. However, existing research has mostly relied on small samples, broad comparisons, and average performance measures, limiting the ability to examine the full range of abilities, including exceptional performance.
This study addresses this gap by examining arithmetic fluency using a dimensional approach across the performance spectrum, while also including individuals with exceptionally high performance.
Method
Phase 1 (Online Testing): All participants complete an online screening designed to assess arithmetic performance and relevant demographic information. Performance on the calculation task is used to qualify participants for Phase 2.
Phase 2 (In-person Assessment and MRI Scanning): We aim to invite 60 participants from Phase 1 (30 exceptional and 30 average performers) for in-person behavioural testing and MRI scanning. Prior to scanning, participants will be asked to complete behavioural questionnaires (e.g. Empathy Quotient, Systemizing Quotient, Autism Spectrum Quotient, Sensory Processing and Repetitive Behaviour Questionnaire).
Community engagement
The study will be informed by consultation with members of the autistic community. We aim to consult with 5-10 people via the ARC database to review the tasks and language used in this study.
Potential impact
A better understanding of the mechanisms behind uneven ability profiles in autism could inform more tailored educational approaches that recognise and support individual strengths, rather than focusing solely on difficulties.
More broadly, the findings may help challenge common social stereotypes about autistic people by highlighting the diversity of abilities across the spectrum. This could improve public understanding of autism as a complex and varied condition, moving beyond oversimplified portrayals often seen in popular culture and supporting more nuanced discussions about individual differences in autism.
Results
We anticipate sharing the results in June 2028. When ready, the key findings will be shared on this page.
Staff:
- Professor Sir Simon Baron-Cohen
- Professor Carrie Allison
- Professor John Suckling
- Professor Jon Simons
- Dr Richard Bethlehem
- Dr Marcin Radecki
- Ramya Warrier
- Huyen Pham
- Matea Balabanovska