Analysis of WISC-III, Stanford-Binet:IV, and academic achievement test scores in children with autism.
Autistic kids’ IQ profiles move with age—watch the verbal-nonverbal gap, then look at attention and writing if achievement lags.
01Research in Context
What this study did
Mayes et al. (2003) looked at IQ and school test scores in kids with autism. They used the WISC-III and Stanford-Binet:IV to map each child’s pattern of strengths and gaps.
The team wanted to see if age or overall IQ level changed the shape of these patterns.
What they found
Younger autistic children scored higher on nonverbal than verbal tasks. As kids got older, the gap flipped for many; higher-IQ students showed new weak spots in attention and writing.
No single profile fit every child—age and IQ level shifted the pattern.
How this fits with other research
McQuaid et al. (2024) extends this work. They show WISC index scores are not statistically separate in autism; working memory bleeds across indices. The finding warns us not to treat small gaps as firm traits.
Estes et al. (2011) also extends the 2003 profiles. They link the same IQ splits to real school marks: many high-functioning 9-year-olds under- or over-achieve versus IQ predictions. Early social skill, not IQ alone, predicts who will read well.
Kalbfleisch et al. (2012) adds another layer. A verbal edge over nonverbal IQ predicts milder parent-rated executive-function problems. Together, these papers show the verbal-nonverbal split is useful, but it must be read alongside age, attention, and social data.
Why it matters
When you test an autistic learner, expect the first score sheet to change. Re-check verbal-nonverbal gaps every couple of years. If the child has high IQ yet struggles in math or writing, probe attention and executive skills next. Use the split as a starting point, not a label.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pull last year’s WISC: if verbal < nonverbal by 12+ points, re-test within 12 months and add a brief attention screener.
02At a glance
03Original abstract
Nonverbal IQs were greater than verbal IQs for young children (3-7 years of age) on the Stanford-Binet:IV (n = 53). However, WISC-III verbal and nonverbal IQs were similar for older children, 6-15 years of age (n = 63). Stanford-Binet:IV profiles were generally consistent for the low-IQ (< 80) and high-IQ (> or = 80) groups, with high scores on visual matching tests (Bead Memory and Quantitative Reasoning). The low- and high-WISC-III IQ groups both performed well relative to IQ on tests of lexical knowledge (Similarities, Information, and Vocabulary), but not on language comprehension and social reasoning (Comprehension). The low-IQ group did best on visuo-motor subtests (Object Assembly and Block Design), but the high-IQ group did not. The high-IQ group had significantly low scores on the Digit Span, Arithmetic, Coding, VMI, and WIAT Written Expression tests, suggesting attention and writing weaknesses.
Journal of autism and developmental disorders, 2003 · doi:10.1023/a:1024462719081