Skip to main content

Numerical Processing Similar in Children, Adults

Study explores how the brain wires itself during development to acquire mathematical skills

Four-year-olds who are still developing numerical abilities show activity in the same brain region during numerical tasks as do math-adept adults, Duke University researchers have found. Their comparative brain scan studies are the first to explore the earliest glimmerings of numerical processing in pre-school children.

The researchers said their findings represent the beginning of a promising new research pathway to explore how the brain wires itself during development to acquire mathematical skills.

The researchers reported in the May 2006 issue of the Public Library of Science Biology that a brain region called the intraparietal sulcus (IPS) is activated when both four-year-olds and adults perceive numerical quantities. The researchers used the analytical technique of functional magnetic resonance imaging (fMRI), in which harmless magnetic fields and radio waves image blood flow in brain regions, which reflects their activity.

Lead author on the paper was Jessica Cantlon, graduate student in the laboratory of co-author Elizabeth Brannon. They are in the Department of Psychological and Brain Sciences and the Center for Cognitive Neuroscience. Senior author on the paper was Kevin Pelphrey, and the other co-author was Elizabeth Carter, both in the Department of Psychological and Brain Sciences. The studies were done in the joint Duke-University of North Carolina Brain Imaging and AnalysisCenter. The research was sponsored by the National Institute of Mental Health, the National Science Foundation, the John Merck Fund and Duke.

"Lots of previous behavioral studies have shown that pre-school children can do basic math tasks before they ever get any formal math training in school," said Cantlon. "They can tell you that a bag of fifteen grapes has more things in it than a bag of five apples, even if they don't know how to verbally count very well. So, it seems like a basic set of math skills are laid down very early in development. And we were interested in whether these early math skills are related to the sophisticated math skills of adults in the brain," she said.

"This study is the first study to use fMRI to study the neural basis of higher-order cognition in children this young," said Cantlon. "This is important because very little is known about how the mind of a young child becomes so sophisticated, especially for mathematics, over a relatively short period of time.

"Our study suggests that the human brain is prepared for basic mathematics at an early age, and that the same neural circuits that perform basic math at an early age continue to process mathematical information over the whole course of development, into adulthood," she said.

"Our ability to scan very young children combined with the design of the experiment were the keys to comparing numerical skills in children and far more sophisticated adults," said Pelphrey.

The experiment involved showing both children and adults a rapid display of objects, for example 32 circles over and over, he said. And when the subjects became accustomed to seeing 32 circles, a display containing 64 circles would appear. The fMRI scans would reveal the brain region activated by this change in number. To ensure that the brain region was not reacting to shape or some other aspect of the stimuli, the researchers would also change the objects to another form, for example a triangle, said Pelphrey.

"The advantage to this approach is that, for the children, we weren't asking them to perform a mathematical operation that was beyond them," said Pelphrey. "We were showing them changes in number and changes in shape that allowed us to find specifically the brain regions involved in numerical perception that weren't connected to the numerical symbols that adults use in math. We clearly found that the IPS responded to number but not to shape in both children and adults."

Also, in behavioral studies, the researchers found that the same children could not verbally count to 64, even though they were capable of discriminating such large numbers when presented in a nonsymbolic way.

Future studies, said the researchers will examine older children, to determine whether there are any detectable age-related differences in numerical processing that might give clues to development of numerical ability. Ultimately, the researchers hope to perform studies in which they follow development of numerical skills in the same children as they grow into adults.

Said Cantlon, "Importantly, this specific brain region is likely only part of a larger network of regions involved in more sophisticated symbolic mathematical skills. So, in such longitudinal studies, we could explore how this region links with language and other numerical-related brain regions."

More broadly, said Pelphrey, study of development of numerical abilities could give fundamental insight into brain development in general.

"The development of numerical processing is a perfect system for using imaging to study the mechanisms involved in cognitive development," he said. "With numerical processing, it's possible to conduct well-controlled experiments using fMRI to explore how changes in the brain relate to changes in cognition."

Keeping notoriously squirmy four-year-olds perfectly still in an fMRI machine was one of the major technical achievements of the study, noted the researchers. The technique: bribery and praise.

The children were first prepared by allowing them to crawl into a fake scanner built by Pelphrey from spare fMRI parts and decorated to look like a spaceship. There, the children watched a cartoon, which would only play when they kept their heads perfectly still. The fake scanner even made the same sounds as a functioning one.

Once the children were comfortable in the fake scanner, they were introduced to the real one. And with Cantlon and Carter offering praise and toy rewards, the children were induced to go through the numerical tests.

"We scanned many more kids than we ended up being able to use in the study," said Cantlon. "Over time, however, we perfected our technique of going into the scanner room with them, instilling confidence in them and giving them praise and rewards."