A new study has discovered a direct link between the number of milliseconds it takes a child’s brain to process the form of a printed word and how well that child understands what they are reading. The finding provides a new way to measure this neural timing in individual children with millisecond precision, a breakthrough that could advance our understanding of how reading skills develop. The research was published in the journal Developmental Cognitive Neuroscience.
The investigation was led by a team of researchers at Stanford University who were interested in the brain changes that support the development of fluent reading from late childhood into early adolescence. During this period, reading often transforms from a slow, effortful task into an automatic and engaging activity. The speed of recognizing individual words is known to be a key element in this transition, but the neural mechanisms behind it are not fully understood.
Previous methods for measuring the brain’s processing speed for words, often using a technique called event-related potentials, have been limited by low reliability when applied to individuals. This makes it difficult to connect brain activity directly to a specific child’s reading ability. The researchers aimed to develop and validate a more precise and stable method to measure this neural timing.
“This study emerged from a unique ‘research practice partnership’ between an innovative Bay Area K-8 school, the Synapse School, and the Stanford Educational Neuroscience Initiative (SENSI),” explained senior author Bruce D. McCandliss, the Pigott Family Graduate School of Education Professor in Educational Neuroscience at Stanford University.
“The collaboration began with a series of roundtable discussions involving teachers and researchers to find synergies between my long-term research goals and the topics that educators found most meaningful. This effort was also informed by my multi-year reflections on the challenges that prevent neuroscience from making a meaningful connection with education.”
“Our first collaborative focus was on how reading changes the brain. We knew we could bring brainwave technology to the school, but a significant limitation of current science is its difficulty in delivering what teachers value most: information that is meaningful at the individual student level. Standard approaches are not yet able to provide this, as their conclusions tend to apply to groups rather than individuals.”
“The teachers stressed the importance of protocols brief enough for students to complete within a single class period,” McCandliss continued. “As a science team, we took this as a design challenge, and created innovative approaches that required only a few minutes of data collection for each measure. The science team also contributed back to the teachers the value that can come from measuring reading skills in units of physics… such as capturing the duration of a specific neural computation with millisecond-level precision.”
To achieve this, the researchers and school staff enabled 68 typically developing children between the ages of 8 and 15 to volunteer for the study during their ordinary school day. Each child participated in a session where their brain activity was recorded using electroencephalography, or EEG, a method that measures electrical signals from the brain through sensors on the scalp.
The children also completed a series of standardized tests to assess their reading abilities, including their speed in reading single words and their fluency and comprehension of sentences.
During the EEG recording, participants viewed rapid streams of four-character stimuli presented at a steady rhythm of precisely three items per second. These stimuli included real words, nonwords made of jumbled letters, and strings of unfamiliar symbols called pseudofonts.
This steady, rhythmic presentation is part of a technique known as Steady-State Visual Evoked Potentials. It is designed to elicit a brain response that follows the same rhythm as the flashing images. The brain produces signals not only at the primary frequency of the stimulus, in this case 3 Hz, but also at its multiples, known as harmonics, such as 6 Hz and 9 Hz.
The researchers analyzed the timing, or phase, of the brainwaves produced at these different frequencies. By examining how the phase of the response changed across the harmonics, they were able to calculate a precise processing delay for each child. This delay, called cortical latency, represents the time it takes for information to travel from the eyes to the brain regions that process visual word forms. This approach allowed for the calculation of a stable neural timing marker for each individual participant.
The researchers found that the measurement of cortical latency was consistent for each child. This neural processing speed remained stable regardless of whether the child was viewing actual words, nonwords, or abstract symbols. This high reliability suggests the method is capturing a fundamental aspect of an individual’s visual processing system.
The researchers also found a strong relationship between cortical latency and the participants’ reading skills and age. Children with faster brain processing speeds, indicated by shorter latencies, tended to have higher scores on tests of both single-word reading efficiency and sentence-level reading comprehension. Similarly, older children generally exhibited shorter latencies than younger children, suggesting that this neural process becomes more efficient with age and experience. These relationships held even after accounting for non-verbal intelligence.
A third finding provided insight into how these processes are connected. The study suggests that the link between rapid neural processing and fluent reading comprehension is largely explained by single-word reading speed. In other words, a faster initial neural response to visual text appears to facilitate quicker and more automatic recognition of individual words. This efficiency at the word level may then free up cognitive resources, allowing the reader to focus on understanding the meaning of entire sentences and passages.
“Your brain is operating at multiple time-scales at the same time,” McCandliss told PsyPost. “You might be aware of how your thoughts or feelings are changing from one moment to the next, and how going to school and learning things allows you to recall new facts. But there’s also many faster time-scales, like the time it takes information to get from your eye to computations that route information to the part of your brain that can recognize words.”
“Remarkably, tiny little differences in neural computation speed for visual words is powerfully tied to your fluency in reading comprehension. As reading improves, this neural timing tends to get faster. Education progressively shapes the speed of this rapid neural process.”
The practical significance of these findings lies in both the strength of the relationship and the reliability of the measurement. The connection between cortical latency and reading fluency suggests that a meaningful portion of the differences seen in children’s reading ability, from a struggling third-grader to a proficient eighth-grader, can be accounted for by millisecond-level variations in neural processing speed.
More importantly, the method used to measure this neural timing proved to be exceptionally stable for individuals. This high reliability is a key advance, as it makes it feasible to track subtle changes in a single child’s brain function over time.
“Because we collaborated with a school in designing and carrying out this study, we know that we can now measure this neural speed incredibly precisely and reliably, like a mechanic timing the sparks in your car’s carburetor, in nearly every school child, within schools, without missing anything more than a single class,” McCandliss explained. “This means kids can see the results of their hard work as learning to read progressively refines this core neural function.”
One of the most surprising outcomes for the researchers was the success of the research-practice partnership itself.
“Collaborating with schools to do brain science was not thought to be viable by most, both on the science side and on the education side,” McCandliss said. “When I told them what I was envisioning, I had people in both science and educational practices look at me in ways that made me think ‘this is pretty crazy,’ or at best ‘that’s a pretty whimsical way to invest your research bets after tenure.’ Looking at the number of scientific discoveries we’ve published, I think the surprise is really how promising these sorts of research-practice partnerships can really be, both for science, and for education.”
A second surprise emerged from the data. The research team was initially uncertain whether their technique would yield a meaningful and stable measurement for any single person. There was a genuine concern that an individual’s brainwave data might be too inconsistent to provide anything other than random noise. “We literally had no idea how well this could work at an individual level,” McCandliss said. Fortunately, the results showed a precise and reliable signal at the individual level “in a way that surpassed what we could have hoped for.”
But there are still some limitations to consider. The study’s design is correlational, which means it identifies a relationship between neural processing speed and reading skill but cannot establish causality. For instance, it remains unclear whether faster processing is a cause of proficient reading, or if, conversely, extensive reading practice is what leads to more efficient neural responses.
“Of course, finding a ‘link’ between measures of a child’s cortical latency and academic achievement in reading is really just the beginning of untangling the dynamics of how this relationship develops,” McCandliss said. “It begs for new studies exploring both how increasingly engaging in reading changes neural timing as well as exploring how differences in neural timing might bias the experience of fluent reading, and how each of these causal pathways may play out over the course of reading development.”
Also, because the study observed children across a wide age range at a single point in time, it is difficult to fully separate the effects of natural brain maturation from the effects of accumulating reading experience.
“Given how this science centers around one of our most vulnerable populations — developing children — it is critical to put this in a developmental context as well as an educational context, which means these are values in flux within a person — they are likely changing as they get more experience in reading and more experience in general,” McCandliss told PsyPost. “The true value of these measures will ultimately be in how we can better understand the way they are changing over time within an individual when given highly valuable learning opportunities.”
Future research could build on these findings in several ways, including by testing the same set of children over several years to observe how their neural processing speed changes in relation to their reading instruction and practice. Such longitudinal studies could help clarify the distinct contributions of age and reading experience to the development of neural efficiency.
The researchers are now conducting additional studies to better understand the nature of this neural signal. They plan to test whether the brain’s processing speed changes in response to different types of stimuli, such as high- versus low-frequency words, or other visual forms like faces and cars. A key goal is to determine if this rapid neural response is truly specific to written language or if it also extends to other complex visual categories that the brain specializes in processing, such as faces.
“One of our next scientific goals is to bring education and developmental neuroscience closer together, which means bringing portable EEG tools into schools, collaborating with the schools to devise precise metrics that can track growth over time in neural computation speed, and ultimately relate time series changes within an individual to variations that matter in their educational experiences,” McCandliss said.
“We also plan to expand these findings into research aimed at the specific challenges facing individuals living with clinically significant neurobiological challenges, such as developmental dyslexia, attention challenges, and autistic spectrum disorder.”
“This study is one part of a suite of papers that resulted from the collaboration between my group at Stanford and the Synapse School,” he added. “I encourage readers to look at them as an interlocking set that shows the true potential of research practice partnerships in advancing developmental cognitive neuroscience research related to education.”
“For example, our group is ecstatic that our latest study was just accepted this month for publication in one of the Nature publishing group journals (npj Science of Learning). This new study shows an actual causal impact of two weeks of teacher’s vocabulary activities on neural responses to words that show up in children’s books less than one in a million times, and brings them to levels of cortical responses equivalent to the highest frequency words we’ve ever tested.”
The current study, “Cortical latency predicts reading fluency from late childhood to early adolescence,” was authored by Fang Wang, Quynh Trang H. Nguyen, Blair Kaneshiro, Anthony M. Norcia, and Bruce D. McCandliss.
