Math abilities in autism spectrum disorder: A meta-analysis.
Autistic learners, on average, score lower on math tests, but individual profiles swing from gifted to delayed.
01Research in Context
What this study did
Tonizzi et al. (2023) pooled 13 earlier studies. They compared math test scores of autistic and typically developing people of all ages.
The team ran a meta-analysis. This means they used statistics to combine every study into one big picture.
What they found
Across the 13 studies, autistic learners scored lower on math tests. The gap was small to medium.
Older kids, lower verbal IQ, and weaker working memory made the gap wider.
How this fits with other research
Chiang et al. (2007) once said math skills in ASD were "mixed." The new meta-analysis sharpens that view. It shows a clear, though modest, average deficit.
Wei et al. (2023) looked only at eighth-grade students. They found autistic students beat peers on visuospatial math yet lagged on word problems. Irene’s wider set confirms the overall lower mean while keeping room for these bright spots.
Leng et al. (2024) traced the trouble back to preschool. Autistic preschoolers were already weaker at telling which set of dots had more. The meta-result lines up: early number sense may seed later gaps.
Titeca et al. (2014) offered hope. They showed that strong preschool subitizing and counting predicted average or above-average first-grade math in ASD. The meta-analysis does not cancel this; it just says the group average still tilts low.
Why it matters
For BCBAs, blanket labels like "autistic kids are bad at math" miss the mark. Expect a small but real risk for delay, then probe early number sense, verbal IQ, and working memory. Build these weak spots into IEP goals while letting visuospatial strengths shine.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add a one-minute magnitude comparison probe to your assessment battery and chart the results.
02At a glance
03Original abstract
BACKGROUND: Studies focusing on math abilities in autism spectrum disorder (ASD) are limited and often provide inconsistent results. AIM: This meta-analysis was conducted to investigate math abilities in people with autism spectrum disorder (ASD) compared to typically developing (TD) participants. METHODS AND PROCEDURES: According with PRISMA guidelines, a systematic search strategy was adopted. First, 4405 records were identified through database searching; then, the title-abstract screening led to the identification of 58 potentially relevant studies and, finally, after the full-text screening, 13 studies were included. OUTCOMES AND RESULTS: Results shows that the group with ASD (n = 533) performed lower than the TD group (n = 525) with a small-to-medium effect (g=0.49). The effect size was not moderated by task-related characteristics. Instead, sample-related characteristics, specifically age, verbal intellectual functioning, and working memory, were significant moderators. CONCLUSIONS AND IMPLICATIONS: This meta-analysis shows that people with ASD have poorer math skills than their TD peers, suggesting the importance of investigating math abilities in autism, taking into account the role of moderating variables.
Research in developmental disabilities, 2023 · doi:10.1016/j.ridd.2023.104559