How autism affects learning can be difficult to see in a child with Autism Level 1, particularly when the child has average or above-average intelligence.
There are two assumptions about autistic children that have never sat particularly well with what I have seen.
The first is that autistic children are often naturally good with numbers. The second is that autism, especially Autism Level 1, is mainly a social difficulty.
Neither idea is completely invented. Some autistic children do have striking mathematical abilities, and social-communication differences are, of course, central to an autism diagnosis. But both ideas become misleading when they are used to describe a developmental profile that is often far more uneven and complex.
For a child without intellectual disability, that unevenness can be particularly easy to miss. A child may be obviously bright, capable in several areas, and still have considerable difficulty learning things that other children seem to acquire with much less deliberate teaching.
I have written before about how this kind of uneven profile can make Autism Level 1 difficult to recognise in small children. The same unevenness matters after diagnosis too, because it affects what we expect the child to learn, when we expect them to learn it, and what we conclude when ordinary teaching does not work.
Uneven does not simply mean behind
One of the easiest mistakes to make after an autism diagnosis is to turn a statistical tendency into a prediction about an individual child.
Motor development is a good example. Research does show, at group level, substantially more gross-motor difficulties among autistic children than among neurotypical children. Balance, coordination and whole-body skills can all be affected. That does not mean every autistic child will be late acquiring every motor skill.
I remember reading a parent’s account after their small child received an Autism Level 1 diagnosis. One of their worries was whether he would learn to ride a bicycle in time. It stayed in my mind because my own experience turned out to be almost the reverse. My child learned to ride at four. At five, he rides a 20-inch bicycle comfortably.
That small example says something important about autism more generally. An uneven developmental profile is not a uniformly delayed profile. The same child can be ahead in one area, entirely ordinary in another, and require substantial help somewhere else. Looking at the strongest ability and assuming everything else should come easily is just as misleading as looking at the weakest ability and lowering expectations across the board.
How Autism Affects Learning Without Intellectual Disability
This distinction becomes more important once formal learning begins. Intelligence and academic learning are related, but they are not interchangeable. Autism research repeatedly finds substantial variation in academic achievement, including among children whose general cognitive ability is not impaired.
A particularly useful study looked at mathematics and reading in 114 autistic children. Rather than finding one characteristic autistic academic profile, researchers found very different groups. One subgroup had relatively poor mathematics compared with reading. Another showed the opposite pattern, with mathematics as a relative strength. Working-memory differences also helped distinguish the groups.
That is much closer to the reality implied by the word spectrum than the familiar picture of the autistic child who loves numbers and excels at mathematics. Some do. Others do not. And even among children who appear good with numbers, there is another question worth asking: what exactly are they good at?
Being good with numbers is not the same as understanding mathematics
Mathematics is not one skill.
Recognising numerals is one skill. Reciting a number sequence is another. Counting objects, remembering arithmetic facts, performing calculations, comparing quantities, understanding mathematical language, solving a problem presented in words, and recognising when a mathematical relationship applies in ordinary life all place somewhat different demands on the child.
A child can therefore show impressive competence with part of mathematics without finding all of mathematics equally easy.
This distinction interests me because the stereotype works in both directions. If an autistic child handles numbers easily, adults may assume mathematical understanding is equally strong. If another autistic child struggles with early mathematics, adults may be surprised because the child does not fit the expected autistic profile.
Neither response is particularly useful. The better question is what the child actually understands, what they can reproduce, what they can apply in a different situation, and where the learning starts to break down. That last part matters because some of the cognitive processes involved in learning can themselves develop unevenly in autism.
Working memory can become a hidden bottleneck
Working memory is one example.
It is the capacity we use to hold information in mind while doing something with it. Even very simple school tasks can place considerable demands on it. A child may need to remember an instruction while carrying it out, hold one quantity in mind while manipulating another, retain an intermediate answer, or keep track of several pieces of information long enough to understand a problem.
Research on autistic preschoolers has linked working memory and other executive-function skills with the rate at which children acquire new mathematics goals. Preschool research more broadly also finds executive-function differences in autism, particularly in areas such as cognitive flexibility, although the profile is far from identical in every child.
What I find particularly important is that these abilities also have developmental trajectories of their own.
A longitudinal study following working memory through the elementary-school years found that autistic children continued to make gains. Their period of greatest developmental plasticity lasted longer, and children who entered school with poorer working memory were more likely to show rapid growth during the later elementary years, when growth among neurotypical peers had begun to level off. The researchers referred to this group as “late bloomers.”
That does not mean every autistic child eventually catches up, or that all learning difficulties disappear with age. It does mean that performance at one point in childhood should not automatically be treated as a final measure of capacity.
Sometimes the child may understand more once the cognitive systems needed to manage the task have developed further. Sometimes the same concept may become accessible much sooner if it is taught in a way that reduces unnecessary demands on those systems.Those are very different possibilities from simply deciding that the child is not good at the subject.
Language can be equally deceptive
Language creates another version of the same problem.
A child’s spoken vocabulary can make their understanding look stronger than it is, while limited speech can make understanding look weaker than it is. Even among autistic preschoolers, studies of language comprehension find considerable individual variation that cannot simply be explained by non-verbal cognitive ability.
And language itself is made of many things. Knowing a word is not quite the same as understanding it in different contexts. Producing a sentence is not the same as following a complicated explanation. Answering a familiar question is not the same as understanding a newly phrased version of it.
This is one area where my own experience has made me particularly cautious about judging what a child knows from what they can say at a given moment.
I will return to language separately, because it deserves more than a few paragraphs. The same is true of mathematics. Both are areas where I eventually found myself thinking less about whether the child could learn something and much more about what had to change in the way the concept was being taught.
Autism is not only a social issue
None of this makes the social dimension of autism unimportant.
Persistent social-communication differences are part of the diagnostic definition. A child would not receive an autism diagnosis without that part of the picture. But describing autism mainly as a social-skills difficulty can still distort what families actually encounter.
For some children, the educational consequences of autism may be substantial even when intellectual ability is average or above average. Difficulties with working memory, flexibility, processing language, self-regulation, generalising what has been learned, or managing several demands at once can affect how efficiently ordinary teaching reaches the child.
That creates a peculiar situation.The child may have enough intellectual ability to learn the material, yet still not learn it reliably through the methods being used. This is where I am uncomfortable with lowering expectations too quickly. Support and accommodation have an obvious place. A child should not be repeatedly pushed through demands that are genuinely beyond their capacity simply because adults prefer the appearance of normality.
But accommodation and development are not opposites.If a child has the capacity to acquire a useful skill, I would rather first ask how that skill can be taught effectively than assume that difficulty acquiring it is evidence that it should not be expected. The distinction matters enormously when the child is still very young and development is moving quickly.
When ordinary learning requires extraordinary input
There is also a practical consequence that is easy to underestimate. When a child does not acquire skills through ordinary exposure or standard teaching, the missing work does not disappear.
Someone still has to notice where understanding is breaking down. Someone has to try another explanation, divide a skill differently, repeat it, put it into context, check whether it has generalised, and sometimes invent a completely different route into the same concept.
In many families, a significant part of that work eventually falls on the parents.
This is part of what I meant when I previously wrote about the amount of developmental work a family can sustainably carry. The work is not always visible from the outside, and it is not necessarily dramatic. Often it consists of thousands of ordinary moments in which something that another child might pick up incidentally has to be noticed, understood and deliberately taught.
For me, that responsibility is not an argument for pessimism. If anything, it is an argument for taking learning seriously. A child who needs more deliberate teaching is not necessarily a child with less eventual capacity. Sometimes we simply do not know the capacity yet.
Looking beyond the autism clichés
The autistic child who is brilliant at mathematics is real. So is the autistic child who struggles with mathematics.
The socially awkward autistic child is real too. So is the autistic child whose social differences are relatively subtle while learning, language, flexibility or processing create far more work in daily life than outsiders would expect.
The mistake is turning any one of those children into the template for all the others.
What I have become much more interested in is the space between ability and performance: what a child appears able to do, what they actually understand, what prevents that understanding from showing itself, and what happens when the method of teaching changes.
That is where I will go next.
There are two areas in particular where I have had reason to think about this very carefully: language, and numbers taught in context. I want to write about them separately, and in considerably more practical detail, because useful teaching methods are only useful to somebody else if the method itself is explained properly.
For now, the point is simpler: with Autism Level 1, normal intelligence does not guarantee an ordinary learning path. And an unusual learning path does not, by itself, tell us where that child will eventually arrive.
Sources
- Chen, Lang, et al. 2019. Quantitative Analysis of Heterogeneity in Academic Achievement of Children With Autism. Clinical Psychological Science.
- Christoforou, Maria, et al. 2023. Executive Function Profiles of Preschool Children With Autism Spectrum Disorder: A Systematic Review.
- Chen, Yanru, and Laudan B. Jahromi. 2025. Self-Regulation and Academic Learning in Preschoolers with Autism Spectrum Disorder: Links to School Engagement and Levels of Autism Characteristics. Journal of Autism and Developmental Disorders.
- Kim, Sohyun An, and Connie Kasari. 2025. Longitudinal Trajectory of Working Memory in School-Aged Children on the Autism Spectrum: Period of High Plasticity and “Late Bloomers.” Journal of Autism and Developmental Disorders.
- Wang, Leah A. L., et al. 2022. Gross Motor Impairment and Its Relation to Social Skills in Autism Spectrum Disorder: A Systematic Review and Two Meta-Analyses. Psychological Bulletin.
- Venker, Courtney E., et al. 2013. Individual Differences in the Real-Time Comprehension of Children With ASD. Autism Research.
- Barone, Rita, et al. 2019. Online Comprehension Across Different Semantic Categories in Preschool Children With Autism Spectrum Disorder. PLOS ONE.
