3 Minutes
They don’t always “not get” numbers. Sometimes the stumbling block is more subtle: a difficulty in updating strategies after an error. That’s the central finding from a new study led by researchers at Stanford, which shifts attention away from pure number sense and toward how young brains respond to mistakes.
Some kids don’t struggle with math because they “don’t get numbers,” but because their brains have trouble adjusting when they make mistakes.
The team, led by Hyesang Chang, asked elementary-age children to perform rapid decisions about which value was larger. On some trials the choices appeared as written numerals; on others they were shown as clusters of dots. Instead of scoring each trial simply as correct or incorrect, the researchers used modeling techniques that tracked how each child’s choices evolved across repeated trials — essentially measuring learning as a dynamic process rather than a static snapshot.
What emerged was a pattern: children who struggled with math were not uniformly wrong across all items. Their performance showed a characteristic instability — they failed to adjust their decision process after mistakes and carried that inconsistency into subsequent trials. In contrast, children with typical math skills tended to update strategies more reliably after an error, leading to steadier improvements in performance.
Prediction of math ability from multiple latent measures of performance in number tasks. Both numerical symbol (symbolic) and dot cluster (nonsymbolic) task formats are shown. Red and blue signify children with typical or atypical math abilities, respectively. *** p < .001; BF = Bayes Factor.
Brains that don’t pivot
Functional brain imaging offered a window into why those behavioral patterns arose. Children with math difficulties showed reduced activation in regions associated with monitoring performance and implementing control — circuits that detect when a plan isn’t working and help shift strategy. These networks include areas commonly implicated in error detection and cognitive control, such as midline and prefrontal systems, which support adapting behavior when feedback signals a need to change.
Crucially, the study’s model-based measures of learning predicted whether a child would fall into the typical or atypical group, and those predictions were tied to the degree of activity in the monitoring networks. In plain terms: weaker neural signals that mark and react to mistakes correlate with weaker gains in numerical decision-making.
That observation reframes how educators and clinicians might approach early math difficulties. If the core issue for some children is not number representation but the ability to revise strategy, interventions that train error awareness, flexible responding, and deliberate strategy switching could complement traditional number-focused tutoring. Hyesang Chang and colleagues plan to expand the model to larger and more diverse groups, including children with other learning differences, to test whether this error-driven mechanism is a common thread across learning challenges.
This study reminds us that learning is a conversation between action and feedback. Sometimes the lesson isn’t about the problem itself but how the brain hears — and acts on — the correction.
Source: scitechdaily
Leave a Comment