Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by differences in how people learn, communicate and interact with others, as well as restrictive or repetitive behaviors. Many past psychology studies have explored the unique features of autism, gathering important insight that has helped to develop educational programs and psychotherapeutic approaches tailored for autistic individuals.
Some research has particularly focused on the relationship between autism and mathematical abilities. Yet findings gathered so far have often been inconsistent, as autistic people can have very different qualities and unique characteristics.
Researchers at the University of Hong Kong and Sun Yat-sen University recently reviewed several past studies focusing on autism to gain a better understanding of how autism relates to mathematical skills. Their paper, published in Nature Human Behavior, suggests that, on average, autistic individuals score lower in mathematical tests than non-autistic individuals, and mathematical abilities vary more greatly among autistic people.
“The fundamental characteristics of math ability in individuals with ASD, specifically proficiency and variability, remain inadequately understood,” wrote Jiaxi Li, Zijun Ke and their colleagues in their paper. “In this systematic review and meta-analysis, we addressed this gap by synthesizing evidence on math ability in autistic individuals relative to the non-autistic population. Searches in multiple databases yielded 66 studies.”
Reviewing past studies and combining their findings
To explore the link between autism and mathematical ability, Li, Ke and their colleagues analyzed the findings of 66 past research studies, while also examining the quality and reliability of the methods used to collect these findings. The team then combined the results of these past research efforts using common statistical methods used to carry out meta-analyses.
“Risk of bias was assessed using an adapted Joanna Briggs Institute checklist. Random-effects meta-analyses used Hedges’ g and natural logarithm of variability ratio (lnVR) as effect sizes,” wrote the authors. “Publication bias was adjusted for using the precision-effect test and precision-effect estimate with standard errors, as well as a three-parameter selection model.”
The researchers also tried to determine whether the mathematical abilities of autistic individuals reported in previous works were influenced by the age of the study participants, their IQ (intelligence quotient) and by implicit biases associated with the study designs. They found that both intelligence and age influenced the variability observed in the mathematical abilities of the autistic individuals that took part in the studies.
“Results show that, compared with the non-autistic population, as represented by standardized norms (mean 100, s.d. 15; 3,051 participants) and typically developing (TD) control groups (2,351 participants), individuals with ASD exhibit significantly lower math scores (ASD versus norms: Hedges’ g = −0.360, 95% confidence interval (CI) −0.605 to −0.114; ASD versus TD: Hedges’ g = −0.696, 95% CI −0.947 to −0.445) and greater variability (ASD versus norms: lnVR 0.159, 95% CI 0.102 to 0.216; ASD versus TD: lnVR 0.298, 95% CI 0.199 to 0.396),” wrote Li, Ke and their colleagues. “Group discrepancies were moderated by intelligence, age or their interactions.”
Towards better education for autistic students
The results of this recent study suggest that most autistic individuals perform worse on mathematical tests than non-autistic individuals. Interestingly, the researchers also found that the discrepancy in the mathematical skills of autistic and non-autistic people has widened over the past few decades.
This highlights the need to improve education for autistic children and young people, taking ASD-related differences into account and better supporting students as they are learning mathematics. In addition, the team identified some methodological issues that might explain the observed inconsistencies among past research findings.
The systematic review and meta-analysis performed by Li, Ke and their colleagues could guide the design of new studies that investigate the link between ASD and mathematical skills. Eventually, their work could also contribute to the introduction of new personalized lower and higher mathematics programs for autistic people.
“Our findings underscore the need for sustained, individualized mathematical education for ASD and investigation of the developmental trajectories of mathematical skills in ASD,” wrote the authors. “Methodological challenges in the field included potential publication bias and insufficient rigor in sample matching.”