Autism & Developmental

Numerical Estimation in Children With Autism.

Aagten-Murphy et al. (2015) · Autism research : official journal of the International Society for Autism Research 2015
★ The Verdict

Autistic children without ID often struggle with number sense and math, so test early and teach specifically.

✓ Read this if BCBAs working with school-age autistic students who have average or higher IQ.
✗ Skip if Clinicians serving only gifted or twice-exceptional autistic learners.

01Research in Context

01

What this study did

Wachob et al. (2015) compared autistic children with typical classmates.

They looked at three skills: guessing which picture had more dots, placing numbers on a blank number line, and a paper math test.

All kids had average or higher IQ; no one had intellectual disability.

02

What they found

The autistic group scored lower on every task.

They guessed dot totals less accurately and placed numbers farther from the correct spot on the line.

Standardized math scores were also lower, busting the myth that autism always brings math gifts.

03

How this fits with other research

Fernández-Cobos et al. (2025) saw the same pattern in younger autistic children without ID. They add that weak visuo-spatial and language skills drive the trouble, not autism severity.

Greene et al. (2019) seems to disagree: their gifted autistic students outperformed typical peers. The clash fades when you notice they studied a high-IQ subgroup, while David et al. looked at the broad cognitively-able range.

Howard et al. (2019) show math scores can jump when you add metacognitive prompts to computer lessons. So the weakness is real, but it can be eased with the right supports.

04

Why it matters

Stop assuming autistic learners will shine in math. Screen early for estimation and number-line errors. Pair these findings with Raúl’s tip to check visuo-spatial and language skills. If gaps show, try Katie’s metacognitive feedback or Zhou’s step-by-step chaining. Targeted teaching beats the wait-and-see approach.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Give a quick dots-estimation and blank-number-line probe; note errors and plan extra visuo-spatial or language supports before the next math lesson.

02At a glance

Intervention
not applicable
Design
quasi experimental
Sample size
64
Population
autism spectrum disorder, neurotypical
Finding
negative

03Original abstract

Number skills are often reported anecdotally and in the mass media as a relative strength for individuals with autism, yet there are remarkably few research studies addressing this issue. This study, therefore, sought to examine autistic children's number estimation skills and whether variation in these skills can explain at least in part strengths and weaknesses in children's mathematical achievement. Thirty-two cognitively able children with autism (range = 8-13 years) and 32 typical children of similar age and ability were administered a standardized test of mathematical achievement and two estimation tasks, one psychophysical nonsymbolic estimation (numerosity discrimination) task and one symbolic estimation (numberline) task. Children with autism performed worse than typical children on the numerosity task, on the numberline task, which required mapping numerical values onto space, and on the test of mathematical achievement. These findings question the widespread belief that mathematical skills are generally enhanced in autism. For both groups of children, variation in performance on the numberline task was also uniquely related to their academic achievement, over and above variation in intellectual ability; better number-to-space mapping skills went hand-in-hand with better arithmetic skills. Future research should further determine the extent and underlying causes of some autistic children's difficulties with regards to number.

Autism research : official journal of the International Society for Autism Research, 2015 · doi:10.1002/aur.1482