Where Maths Difficulty Really Lives: The Role of Attention

What the research found

A child who cannot learn arithmetic invites two explanations. The first places the difficulty in number itself, in an unreliable sense of quantity that makes comparing amounts, counting, and recognising how many objects are on a table harder than they should be. The second places it in the general equipment a child brings to every subject: how quickly the mind takes in what it sees, how much it can keep available while working on something else, how steadily it stays with a task. Researchers call these the domain-specific and domain-general accounts, and they organise the field.

Francesca Agostini and Maria Casagrande, working with Pierluigi Zoccolotti at Sapienza University of Rome, asked what the published evidence for the second account amounts to. Their systematic review appeared in the journal Brain Sciences in 2022 and assembled 46 studies, 31 comparing groups at a single point and 15 following children across years of school. Between them the studies tested 8,506 primary-age children between 1980 and 2021, and each measured intelligence as well as mathematics, so that a general intellectual deficit could be excluded as the source of the maths result.

Attention produced the most consistent finding. All nine studies that assessed it found children with mathematical difficulties performing below their classmates. Processing speed showed a difference in 17 of 22 studies, verbal working memory in 16 of 21, visuospatial working memory in 9 of 14, inhibition in 6 of 8, and cognitive flexibility in 4 of 7.

These children responded as quickly as their peers on simple reaction time tests, where the whole requirement is to press a button when something appears, and were slower on visual search tasks, where a stimulus must be examined and discriminated first. The Corsi Block test, a standard measure of visuospatial memory, turned up no group difference in any study that used it, while other measures of the same ability turned up differences repeatedly. Within a single domain, the choice of instrument decided the result.

No pooling was possible, because the tests varied so widely that a meta-analysis could not be run. The tallies record how many studies reached statistical significance and say nothing about the size of any difference. Entry to a mathematical difficulties group ranged from the 10th percentile in some studies to the 35th in others, so the children being compared were not the same children from one paper to the next. Across all of that variation, attention was the single domain in which every study agreed.

What it means for you and your child

The general skills are separable and measurable, which matters because two children can reach the same maths score by different routes, and the review’s tallies are an attempt to name the routes. A child who cannot retrieve a times table and a child who loses the thread halfway through a long division are recorded identically on most standardised tests, and the second child’s difficulty shows itself in tasks that have nothing to do with number.

Multi-step arithmetic asks a child to keep an interim result available while performing a further operation on other digits, and to notice when the procedure has to change from one rule to another. Those are the abilities the review found disturbed most often, which supplies a plain explanation for what families describe: mathematics is the first subject to go when a child is tired, rushed, or working in a noisy room.

Assessment of general cognitive skills is already recommended in the diagnostic guidance for dyscalculia, though the authors note there is no agreement on which skills deserve it. Their proposal is that speed, working memory, attention and flexibility be assessed alongside number skills; which of them earns a place there is the part the field has not agreed. Studies of intervention were excluded from the review by design, so the further suggestion that a child whose maths difficulty follows from weak attention might be helped by addressing attention first is an idea the authors advance, and nothing in these 46 studies tests it.

The review cannot order cause and effect, so whether slow processing produces the arithmetic difficulty, or years of arithmetic failure produce a child who disengages from the tasks that measure processing, is undetermined by this design. Nor can it weigh the two explanations against each other, since the studies it gathered were selected for measuring general skills, with the sense of number outside their remit. The review even disagrees with itself on verbal short-term memory, where its summary table counts nine of twelve studies finding a difference and its discussion concludes that short-term memory is unaffected. Which explanation fits your child is a question the field has not yet built the study to answer, and anyone who answers it with confidence is running ahead of the evidence.