The IQ Gap Closed After Correction

What the research found

This study asked a narrow, useful question. If you give one broad test battery to autistic and non-autistic children who are otherwise closely matched, where do their profiles actually differ? The researchers used the Intelligence and Development Scales–2, or IDS-2, a standardised battery that measures both intelligence and a wider set of developmental functions such as motor skills, language, and social-emotional understanding. It is designed for people aged 5 to 20.

The sample was small and carefully constructed. Forty-three autistic children and adolescents were each paired with a non-autistic child matched on age, sex, and mother’s education level, giving 86 participants in all. They were drawn from German-speaking regions and ranged across the school years. Matching in this way is a deliberate attempt to compare like with like, so that any differences are less likely to be explained by age or background alone.

The headline result concerns where the two groups diverged. The clearest differences were not in intelligence but in physical and language skills. Autistic participants scored significantly lower in psychomotor skills, the umbrella term for coordinated movement. The gap in gross motor skills was large, with a t-value of 5.30 and a p-value below .001, and effect sizes ranging up to d = 1.82. Language skills showed a similar pattern, with receptive language, the ability to understand spoken language, producing effect sizes as large as d = 1.63. These are substantial differences, not statistical hair-splitting.

Intelligence told a subtler story. On the raw comparison, autistic participants scored a few points lower. Their Full-Scale IQ averaged 91.58 against 97.63 for the matched group, a modest gap with an effect size of d = 0.35 that did not reach significance even before adjustment. The Profile IQ difference was slightly larger, at d = 0.44. But here the study did something honest. Because it ran many comparisons at once, it applied a statistical correction to guard against chance findings. After that correction, none of the intelligence differences remained significant. Two memory measures, auditory and visuospatial short-term memory, also faded once the same correction was applied.

There was one further pattern worth noting, and worth holding loosely. The differences appeared concentrated in the younger children rather than the adolescents. Among the older participants, the researchers found no significant group differences across the cognitive and developmental measures at all. This hints at a developmental story, but with so few participants split into age bands, it is a hint and not a conclusion.

What it means for your family

The first thing to take from this is a note of caution about the word “cognitive.” When people picture an autistic child’s cognitive profile, they often think first of intelligence. This study, on this battery, found that intelligence was where autistic and non-autistic children looked most alike. Once the researchers corrected for testing many things at once, the IQ gap essentially closed. The real distances lay in movement and in understanding and using language. Those are the areas where support may matter most, and they are easy to overlook if attention fixes only on IQ scores.

It also helps to be clear about what a study of this size can and cannot show. Eighty-six children is respectable for this kind of detailed testing, and the matching is a genuine strength. But it is still a group-level portrait. The authors say so plainly: their analysis describes averages, and averages do not describe your individual child. A given autistic child may have strong motor skills and find language effortless. The profile here is a tendency across a group, not a template for any one person.

Two further limitations, both acknowledged by the authors, are worth carrying with you. First, the diagnoses came from various clinical services, so the researchers could not confirm they were made to a single consistent standard, and they had no information on symptom severity or on what therapies each child had received. Roughly half the autistic sample had what was then called Asperger’s syndrome, meaning average or above-average intelligence, so this is not a picture of the whole spectrum. Second, and importantly, nothing here establishes cause. The study is a snapshot in time. It cannot tell you why the differences exist, whether they narrow as children grow, or whether intervention drives the smaller gaps seen in the adolescents. That the older children differed less is intriguing, but it could reflect maturation, support received, the particular children who happened to take part, or simple chance in a small subgroup.

For a parent, the practical reading is measured. If your child is autistic, this is one reason not to assume that a lower score on a movement or language task reflects lower general ability. The two can come apart, and this study is evidence that they often do. It is also a reminder to read any single assessment as one data point among many, gathered on one day, by one battery. The value of work like this is not that it settles anything, but that it sharpens the questions worth asking about where a particular child is thriving and where a little more support would help.