Cognitive Neuroscience of Learning: Implications for Education

Websites

Session 3

The Number Race is free software developed by cognitive scientist Stanislas Dehaene's group to prevent or remediate dyscalculia or promote math skills among kindergartners. Evaluations of the software indicate that it promotes skill in numerical comparison tasks and numeracy.

Panamath, a program developed at Johns Hopkins University, measures the approximate number sense by asking users to quickly make numerical judgments about groups of colored dots. The program is available as downloadable software or as an online test.

More information about the philosophy and practice of JUMPmath's curriculum is found on its website, which also sells workbooks and teachers' guides.

Video: Infants and primates possess an "approximate number sense." In a research video from speaker Elizabeth Brannon's lab, an infant is fascinated by the stimulus on the viewer's left, which alternates between showing 8 and 16 dots, and ignores the stimulus on the viewer's right, which shows a constant number of dots. This preference suggests the baby can tell the difference, reflecting rudimentary number sense. Lemurs also possess a basic grasp of numbers; another video shows a research animal correctly putting pairs of stimuli in numerical sequence.

Sessions 5 & 6

The University of Oregon Brain Development Lab developed a video for a general audience that explains how the human brain grows and develops, using evidence-based scientific findings to explore cognition and learning.

Judith Willis's website describes her educational philosophy, with links to her articles and books.

The Learning Resource Network will offer evidence-based information for parents and educators to learn about childhood education and learning.

The mission and rationale behind the Ultimate Block Party is described at their site, along with resources for more information about play, creativity and learning.

Journal Articles

Session 1

Carlson SM, Davis AC, Leach JG. Less is more: executive function and symbolic representation in preschool children. Psychol. Sci. 2005;16(8):609-616.

McGrath LM, Pennington BF, Shanahan MA, et al. A multiple deficit model of reading disability and attention-deficit/hyperactivity disorder: searching for shared cognitive deficits. J. Child Psychol. Psychiatry 2011;52(5):547-557.

Raver CC, Jones SM, Li-Grining C, et al. CSRP's Impact on low-income preschoolers' preacademic skills: self-regulation as a mediating mechanism. Child Dev. 2011;82(1):362-378.

Willcutt EG, Betjemann RS, McGrath LM, et al. Etiology and neuropsychology of comorbidity between RD and ADHD: the case for multiple-deficit models. Cortex 2010;46(10):1345-1361.

Session 2

Goswami U. A temporal sampling framework for developmental dyslexia. Trends Cogn. Sci. (Regul. Ed.) 2011;15(1):3-10.

Goswami U, Sz?cs D. Educational neuroscience: Developmental mechanisms — towards a conceptual framework. Neuroimage 2011;57(3):651-658.

Maurer U, Bucher K, Brem S, et al. Neurophysiology in preschool improves behavioral prediction of reading ability throughout primary school. Biol. Psychiatry 2009;66(4):341-348.

McCandliss BD. Educational neuroscience: the early years. Proc. Natl. Acad. Sci. USA 2010;107(18):8049-8050.

Yoncheva YN, Zevin JD, Maurer U, McCandliss BD. Auditory selective attention to speech modulates activity in the visual word form area. Cereb. Cortex 2010;20(3):622-632.

Session 3

Cantlon JF, Brannon EM. Basic math in monkeys and college students. PLoS Biol. 2007;5(12):e328.

Halberda J, Mazzocco MMM, Feigenson L. Individual differences in non-verbal number acuity correlate with maths achievement. Nature 2008;455(7213):665-668.

Holloway ID, Ansari D. Mapping numerical magnitudes onto symbols: the numerical distance effect and individual differences in children's mathematics achievement. J. Exp. Child Psychol. 2009;103(1):17-29.

Libertus ME, Brannon EM. Behavioral and neural basis of number sense in infancy. Curr. Dir. Psychol. Sci. 2009;18(6):346-351.

Mazzocco MMM, Feigenson L, Halberda J. Preschoolers' precision of the approximate number system predicts later school mathematics performance. PLoS ONE 2011;6(9):e23749.

Session 4

Baijal S, Jha AP, Kiyonaga A, Singh R, Srinivasan N. The influence of concentrative meditation training on the development of attention networks during early adolescence. Front Psychol. 2011;2:153.

Bavelier D, Green CS, Dye MWG. Children, wired: for better and for worse. Neuron 2010;67(5):692-701.

Diamond A, Lee K. Interventions shown to aid executive function development in children 4 to 12 years old. Science 2011;333(6045):959-964.

Dumontheil I, Roggeman C, Ziermans T, et al. Influence of the COMT genotype on working memory and brain activity changes during development. Biol. Psychiatry 2011;70(3):222-229.

Dye MWG, Green CS, Bavelier D. Increasing speed of processing with action video games. Curr. Dir. Psychol. Sci. 2009;18(6):321-326.

Jha AP, Krompinger J, Baime MJ. Mindfulness training modifies subsystems of attention. Cogn. Affect Behav. Neurosci. 2007;7(2):109-119.

Klingberg T. Training and plasticity of working memory. Trends Cogn. Sci. (Regul. Ed.) 2010;14(7):317-324.

Mackey AP, Hill SS, Stone SI, Bunge SA. Differential effects of reasoning and speed training in children. Dev. Sci. 2011;14(3):582-590.

Söderqvist S, McNab F, Peyrard-Janvid M, et al. The SNAP25 gene is linked to working memory capacity and maturation of the posterior cingulate cortex during childhood. Biol. Psychiatry. 2010;68(12):1120-1125.

Sessions 5 & 6

Carew TJ, Magsamen SH. Neuroscience and education: an ideal partnership for producing evidence-based solutions to Guide 21(st) Century Learning. Neuron 2010;67(5):685-688.

Deslauriers L, Schelew E, Wieman C. Improved learning in a large-enrollment physics class. Science 2011;332(6031):862-864.

Subotnik R. F., Olszewski-Kubilius P., F.C. Worrell. 2011. Rethinking giftedness and gifted education: a proposed direction forward based on psychological science. Psychological Science in the Public Interest 12: 3-54.

Sponsors

Presented by

Silver Sponsor

Aetna Foundation

Bronze Sponsor

Lindamood-Bell Learning Processes

Academy Friend

Great Leaps Reading and Math Fluency

LearningRx

Scientific Learning Corporation

Grant Support

National Science Foundation

Cognitive neuroscience is providing new insights into how the brain learns. It can identify the specific neurological deficits that underlie learning disorders and suggest new ways to correct deficits. In the future, it may also suggest new ways to accelerate the learning process by providing a window into why certain techniques work and into which types of education work best for whom.

In studies of reading disabilities, neurocognitive research is identifying underlying deficits that point the way toward teaching techniques that could ameliorate such problems. As a child begins spelling out words, directing his or her attention to phonology—how words sound—improves literacy acquisition and has a distinct effect on brain dynamics, said organizer Bruce McCandliss of Vanderbilt University. Other insights offered by neuroscience include a new understanding of the role of perceptual deficits in dyslexia and observations of the neural correlates of reading acquisition.

While mathematical skill may seem less natural than language learning, cognitive science demonstrates that babies and even other primates possess an "approximate number system," the foundation of mathematical ability. Toddlers with a high-functioning approximate number system typically grow into school-aged children who are good at formal arithmetic. How best to cultivate this ability—and to what degree it can be improved—is an open question. Brain activity studies suggest that as math skills improve, the brain becomes more efficient at allocating resources to computation.

Executive function, the "wise old owl" of the brain, includes cognitive flexibility, planning, and controlling attention.

Cognitive neuroscience also probes the skills known as "executive function," capacities such as attention, working memory, and self-regulation that enable learning. Keynote speaker Goldie Hawn, whose Hawn Foundation implements a school curriculum that teaches social and emotional learning skills, referred to executive function as the "wise old owl" of the brain. This clever shorthand refers to the many related faculties of executive function, including cognitive flexibility, planning, delayed gratification, and the ability to control attention. Researchers have recently begun to identify the neural correlates of such capacities, permitting a brain-based understanding of what these skills are and of how they develop. When measured in young children, self-regulatory skills predict an impressive array of adult outcomes, including academic achievement, high school and college graduation rates, and even obesity risk and frequency of physical injuries.

What neuroscience can do

A brain-based understanding of learning could improve educational techniques, particularly for disadvantaged or learning-disabled children. Neuroscientific methods might identify the fundamental deficits that underlie a problem like dyslexia, identify which students will benefit most from an intervention, target students who are at risk of falling behind, or offer an early indication of whether an approach is actually helping. Cognitive neuroscience also offers new ways to test interventions such as art and music training, mindfulness-based programs, physical activity, and social gaming programs. The MindUP curriculum developed by the Hawn Foundation, for example, teaches small children about the brain, and incorporates techniques that promote behavioral and emotional self-control, reduce stress, improve concentration, and increase empathy. Many of these promising curricula have yet to be rigorously tested.

Neuroscience might identify fundamental deficits, target students at risk, or suggest whether an approach is helping.

At the policy level, neuroscience can provide fresh insights into unresolved debates among educators, such as whether so-called "authentic practice"—teaching an entire skill at once—is more successful than breaking tasks into many microskills to be learned incrementally. By monitoring how the brain changes during learning, neuroscience can evaluate competing theories about how best to help minds learn. These opportunities are particularly compelling in light of persistent academic achievement gaps between minority and white students that have not been addressed by conventional educational techniques.

Introducing neuroscience into educational policy and practice is a work in progress. Interventions based on the best science may prove impractical in the classroom or may not be implemented effectively. More broadly, educators and neuroscientists don't speak the same language. Cognitive neuroscientists who study learning are sometimes disappointed to discover that curricula are inspired by ideas that have been discredited by brain science. Educators and policymakers may not see the relevance of the perspective offered by neuroscience or may fail to understand why cognitive science could offer special expertise.

In convening this conference, the organizers hoped to encourage researchers to identify the most important scientific questions and to set a research agenda to shape future exploration. At the same time, some established findings might already be integrated into educational practice, and another focus of the conference was to formulate a plan for meaningful policy change. Influencing policy, speakers agreed, would require intensive collaborations between cognitive scientists, educators, and policymakers.

"It's clear to me that this is a very exciting, rapidly developing field with enormous potential applications," said Carl E. Wieman, Associate Director for Science in the White House Office of Science and Technology Policy, in his keynote address. Neuroscientists interested in education can play two important roles, he suggested: Advancing the state of knowledge of how learning happens, and changing the discussion so that learning is understood as a complex biological process. In this new view of education, learning is what reshapes and redefines the brain.

Speakers:
Stephanie M. Carlson, University of Minnesota
Clancy Blair, New York University
Dale C. Farran, Vanderbilt University
Bruce F. Pennington, University of Denver

Highlights

• Executive function (EF) includes such skills as attentional flexibility, working memory, and behavioral control.
• Evaluating techniques to improve EF requires measuring it in children under 3 years of age.
• EF in early childhood predicts many adult outcomes.
• In a randomized clinical trial, the Tools of Mind curriculum did not improve EF or academic achievement.
• Developmental disorders such as dyslexia often involve comorbidities.

Understanding executive function

Executive function (EF) is an umbrella term covering mental faculties from cognitive flexibility to inhibitory control to working memory, said Stephanie Carlson of the University of Minnesota in her overview. It is manifested in attention control, rule use, planning, and gratification delay. The degree of EF in childhood predicts a wide range of academic and social outcomes in adulthood, from criminal behavior to obesity to scholarly achievement. Developing EF involves a transition from reflexive to reflective behavior, as an impulsive toddler grows into a preschooler who can begin to execute goal-directed actions and then matures into a school-age child who can reflect on his or her own actions and cognition. Developmentally, EF improves rapidly in children between ages three and four, continues to change through the course of childhood, and may be malleable into the adult years.

Developmental psychologists have become increasingly interested in childhood executive function in recent years. (Image courtesy of Stephanie Carlson)

Reflection is key to EF, Carlson proposed, because it provides a delay between stimulus and response that permits behavior to be evaluated before it is enacted. In her research she uses a technique that encourages children to engage in symbolic play—in this case imaginative storytelling—which stimulates flexible and abstract thinking, to demonstrate that it is possible to measure and to begin promoting EF in children as young as two and half years old.

Self-regulation, which includes the processes of executive function, is strongly related to school readiness and underlies learning abilities, said Clancy Blair of New York University. Many stress-related factors shape EF, including home environment and parenting quality. To measure whether EF in preschoolers predicts academic ability in kindergarten, his group has developed tasks that can be used to evaluate children as young as three years old. Because this measurement effort is still underway, Blair discussed data from a longitudinal survey of poor families called the Family Life Project. These data suggest that improvements in the home between ages three and five predict improvements in EF, an effect mediated by stress responses. Techniques to improve executive function should be experimentally evaluated, suggested Blair: The techniques "can help tell us what the best activities and ways are to promote school readiness."

Techniques to improve executive function should be experimentally evaluated.

The results of such an evaluation were presented (via Skype) by Dale Farran of Vanderbilt University, who dialed in to summarize negative findings from a randomized clinical trial of the "Tools of the Mind" curriculum. Tools of Mind was developed by Elena Bodrova and Deborah Leong in part to promote executive function skills in children, but, Farran said, "Aspects of the curriculum appear to have the opposite effect of that intended by the developers." As implemented with about 800 children aged 4 and 5 years old from low-income families, the program showed no effects on literacy, language, or math achievement, nor did it improve executive function measures as assessed by attention shifting, inhibitory control, and memory tasks.

The data challenge two assumptions about promoting EF. "Self talk," in which children are encouraged to talk through their actions, was negatively correlated with both EF and achievement gain. Make-believe play was also negatively related to EF gains. These activities should be better understood before they are advocated as ways to improve EF, suggested Farran.

When learning goes wrong

Developmental disorders occur in many shades of grey and often involve comorbidities, said Bruce F. Pennington of the University of Denver in his overview of learning disabilities. Between 25% and 40% of children with symptoms of dyslexia (or reading disability, RD), defined as poor spelling and difficulty in accurately or fluidly recognizing words, also show evidence of ADHD, which involves hyperactivity, impulsivity, and inattention. Speech sound disorder (SSD), a delay in the acquisition of speech sounds that may involve both phonological and motor difficulties, seems to predispose children to develop dyslexia. However, the cutoff for what qualifies as a disorder is based on behavioral criteria rather than on etiology. He said, "Definitions are little bit arbitrary because of that, but the great advantage is that we can study these symptom dimensions continuously," and that studies can include typically-developing children as well as those with disorders.

Attending to comorbidities is important because one disorder may influence the course and treatment of another, and co-occurrence may provide a window into the developmental causes of both disabilities. The overlap in particular individuals suggests that the genetic risk loci associated with RD likely play a role in SSD as well. Studies that investigate the combined effects of genes and environment suggest a "bio-ecological" interaction in which the genetic contribution to reading disorders may be more easily detected in families with high educational status. "If your parents are doing everything they can to help you learn to read, and you still don't read, the reason is more likely to be genetic than environmental," Farran said. Reading disability is also frequently comorbid with and shares genetic risk factors with ADHD and language impairment.

Speakers:
Mark S. Seidenberg, University of Wisconsin
Bruce D. McCandliss, Vanderbilt University
Kenneth R. Pugh, Haskins Laboratories
Usha Claire Goswami, University of Cambridge

Highlights

• African American children may face a more complex task when learning to read because of dialect differences.
• Top-down processes such as attention are involved in the earliest stages of reading.
• Educators can shape brain plasticity by directing attention.
• Learning to read calls upon auditory and motor patterns shaped by speech acquisition.
• Dyslexia may result from a problem in brain activity synchronization.

The brain begins to read

The so-called "achievement gap" in academic performance between children of minority groups such as African-American children and others has remained consistent for decades, said Mark S. Seidenberg of the University of Wisconsin. Only part of the shortfall can be explained by poverty, and the gap actually gets larger as schooling progresses. One cause might be differences in spoken language between standard American English and African-American English, a dialect in which speakers may pronounce as many as 30% of words differently. Students more familiar with this dialect must perform a kind of simultaneous translation or "code-switching" in their minds as they learn to read in order to form connections between spoken and spelled words.

The achievement gap between African-American and white students has remained consistent for decades despite efforts to improve the disparity. (Image courtesy of Mark Seidenberg)

In his experiments, words that are pronounced differently in this dialect such as "gold" or "ask" result in longer latencies in word naming for both children and adults, suggesting additional cognitive processing is taking place. When a computer model was taught to pronounce words, inputting data from both African-American English and standard English pronunciations led to poorer performance, an effect reduced by providing additional context within the sentences for the words. African-American students might learn to read more readily if provided with greater context for differentiating between these dialects.

Learning to read requires tremendous neural plasticity, as regions specialized for visual attention make new connections with areas involved in memory, speech production, and hearing comprehension. Instructors can influence this neural change by selectively directing the attention of beginning readers, said Bruce McCandliss of Vanderbilt University. In beginning readers, the shift to processing in the left hemisphere or "left lateralization" of a neural response that occurs half a second after an auditory phonological stimulus change (such as "ba" following a string of "da"s) was highly predictive of reading achievement in second grade, suggesting a way to identify at an early time point children at risk of dyslexia.

Instructors can influence neural change by directing attention in beginning readers.

Teachers shape how the brain changes during reading acquisition; phonics curricula, which focus on sub-syllabic sounds, have "profound outcomes" on literacy, McCandliss said. In his experiments, auditory stimuli that blend a word and a series of tones were used to explore how the brain tunes into phonemic information. People asked to pay attention to the meaning of the words and not the tone activated the left-lateralized brain circuit as well as visual regions. "You can focus your mind in order to process these stimuli differently," he said, demonstrating that instruction can change which brain mechanisms are recruited during learning.

The neuroscience of reading disabilities

Cognitive neuroscience can provide a "deeper account" of individual differences in development and in responses to interventions, said Kenneth Pugh of Haskins Laboratories. Some findings point to left hemispheric structural abnormalities associated with reading disability. Proficient readers activate a well-integrated left hemisphere posterior network in the brain, and increasing skill is associated with greater left hemispheric specialization. Those with reading disabilities fail to engage left hemisphere tempoparietal and occipitotemporal regions coherently, and show more activity in the right hemisphere. Effective treatments partially normalize this left hemispheric pattern. "What we know about the brain basis of atypical reading is still largely descriptive," cautioned Pugh; identifying causal mechanisms will require integrating studies in genetics, behavioral studies, and neuroimaging.

Neuroimaging studies suggest dyslexic readers (bottom) do not enlist left hemispheric tempoparietal networks. (Image courtesy of Kenneth Pugh)

Focusing on network activity abnormalities in dyslexia, Usha Claire Goswami of the University of Cambridge described how sensory impairments affect the process of learning to read. In reading, the brain makes use of auditory and motor processing patterns established during the acquisition of language. Brains that are good at reading are phonologically adept as well, and the ability to rhyme words or count syllables predicts later reading skill. The fundamental deficit in dyslexia may actually involve the brain's ability to sample and encode speech information at specific frequencies.

The ability to rhyme words or count syllables predicts later reading skill.

How the brain interprets aspects of the "speech envelope" such as syllable onsets and rise times—the speed at which the sound reaches its peak amplitude—are related to how it learns to read. Failure to process the speech envelope may make it difficult for the brain to engage attention and to synchronize network activity in such a way that generates perceptual binding—the synthesis of streams of sensory information into a coherent impression. The dyslexic brain may be unable to respond to auditory stimuli by synching itself (entraining) to the rhythms of speech. One potential corrective, if the problem is identified in young children, might be early rhythmic inputs such as nursery rhymes, singing, and other forms of musical speech.

Magnetoencephalography of high-functioning dyslexic college students showed abnormal phase locking at 2 Hz (second column from left), roughly the rate of syllables in spoken speech. (Image courtesy of Usha Goswami)

Speakers:
Elizabeth M. Brannon, Duke University
Lisa Feigenson, Johns Hopkins University
Edward M. Hubbard, Vanderbilt University
Daniel Ansari, University of Western Ontario
Daniel L. Schwartz, Stanford University
John Mighton, JUMP Math

Highlights

• Humans and primates share an innate "approximate number system" (ANS) for estimating number.
• ANS acuity is linked to and predicts school math ability.
• As children learn math, their brains begin using the same circuits to process numerals and "real-life" quantities.
• Increasing math fluency is correlated with reduced activity in an anterior region, perhaps due to greater efficiency.
• The brain employs perceptual information to comprehend abstract mathematical functions.
• Neuroscientific findings can help resolve ongoing debates in educational policy.
• A math curriculum emphasizing incremental learning dramatically improved math scores in some classes.

The innate number system

Sophisticated mathematical reasoning requires a human brain, as well as extensive schooling. But, explained Elizabeth M. Brannon of Duke University, babies and even animals possess an "approximate number system," (ANS)—an innate capacity to make rough numerical estimates. Monkeys, for example, can add and subtract, although their estimates are not precise. Six-month-old babies recognize the difference between 8 and 16 dots.

How does this ability relate to mastering the integer system that we learn in classrooms? Lisa Feigenson of Johns Hopkins University has investigated this relationship and its educational implications. ANS accuracy can be measured with a stimulus that includes many blue dots and many yellow dots. Subjects are allowed only 200 milliseconds to decide which color has more dots, forcing them to estimate. Variability between individuals' estimates is huge, and evidence suggests ANS performance generally predicts success in symbolic math. Children's performance at age three on the dot task predicted their ability at age six on a test of symbolic mathematical ability. A longitudinal study underway now indicates a causal link between these abilities, and suggests that ANS training in four-year-olds can improve math skills later on.

Training the approximate number system in four-year-olds might improve math skills later on.

During early schooling—kindergarten through third grade—portions of parietal and frontal regions of the brain are retrained to recognize the relationships between quantities and mathematical symbols and to manipulate those symbols, said Edward M. Hubbard of Vanderbilt University. His group identified three regions of cortex that are responsive to digits, an activity pattern that is modulated by experience. In neuroimaging studies comparing kindergarteners to older children, training in arithmetic causes the brain to activate the same frontal and parietal circuitry to process quantities and formal numerals.

Compared to adults, kindergarteners show more activity in anterior regions, suggesting that as age and skill advance, this region is less important in math processing. "It looks like we're seeing a nice neural correlate of increasing math fluency at the level of brain efficiency," said Hubbard. Understanding how these systems are fine-tuned may suggest ways to improve math skill in students.

How the brain learns math

The parietal cortex becomes increasingly involved in number processing with age, but the functional significance of that change, and its utility for measurement and diagnosis, are open questions, said Daniel Ansari of the University of Western Ontario. In number comparison tests, subjects quickly identify the larger of a pair of numerals. Reaction times are consistently slower for numbers that are close together, a phenomenon called the "numerical distance effect" or ratio effect. This effect is more pronounced in children with poorer math scores. Neuroimaging in children identifies the left intraparietal sulcus (IPS) as involved in this effect: Those individuals with the strongest change in IPS activity during a number comparison test showed the smallest ratio-related decrement in accuracy and response times.

Children who show the smallest change in inferior parietal sulcus activity when asked to judge numbers that are very similar also score highest in mathematical fluency (far left of chart). (Image courtesy of Daniel Ansari)

Daniel L. Schwartz of Stanford University described the role of perceptual systems in thinking about negative numbers and zero—a highly abstract task. When asked to add small negative and positive numbers, fourth graders ultimately performed better when they used a "symmetry" strategy by creating a mental image of positive and negative numbers reflected around the midpoint of zero. Adults asked to decide whether one number was the midpoint between two others activated regions of the brain involved in processing symmetry. "We seem to make sense of higher order concepts with perceptual faculties," Schwartz said. He is investigating whether symmetry-focused instruction could be useful in the classroom, and if such instruction integrates activity between brain regions associated with symmetry and with mathematical analysis.

Brain-based studies might help resolve theoretical knots in education.

Schwartz, describing himself as an "educator moving into neuroscience," proposed that brain-based studies might help resolve "theoretical knots" in education policy, such as the debate about whether to prioritize hands-on learning, which foregrounds sensory information, or to emphasize the automaticity of math skills. Neurocognitive research could also generate theoretical guidelines that educators could use to shape day-to-day classroom practice.

A math program that relies on "guided discovery" (hands-on learning with intensive feedback) combined with Socratic-style inquiry, personalized attention, and incremental learning was described by John Mighton of JumpMath, a Canadian educational non-profit. In one Toronto class, 5th graders who scored collectively at the 55th percentile on a standardized test advanced to the 98th percentile after a year of this curriculum. Similarly strong results were measured in classrooms in a poor district of London. Innumeracy is a "deep spiritual loss," said Mighton, who dismissed the belief that some children just aren't good at math: "Kids love to solve puzzles," he said. "We shouldn't drum that out of them as they move through school." A randomized controlled study of the curriculum underway by the University of Toronto is not yet complete, but early analysis indicates that students taught with the JUMP curriculum show twice the rate of improvement than do students experiencing typical classroom instruction.

Old questions, new media

A panel discussion featuring Schwartz, Daphne Bavelier of the University of Rochester and Torkel Klingberg of the Karolinska Institute considered technology and social media in education. Gaming can be a powerful tool to teach skills, with advantages such as precise scalability (games can be calibrated to be exactly as difficult as they need to be to promote learning) and repetition, since a fun game will be repeated many more times than a neuropsychological task. Learners can develop a more intuitive, implicit sense of an abstract concept such as statistical distributions through gaming, said Schwartz. Aspects of social media can also promote learning; children asked to teach a virtual avatar who represented a fictional character worked harder and longer at the task than those who were assigned a virtual representation of themselves, he said. With social media, "you can integrate personalized training without losing social interactions," said Bavelier.

An open question with new-media–based training, as with any learning, is generalizability: Our brains get very good at the specific skills we practice, but expertise in one domain may not translate to another. General skills, such as knowing what to pay attention to, may be more important than the specific skills gained through any practice, whether it is gaming, puzzles, or chess. Neuroscience might help define the optimal task variance for skill transfer—some evidence suggests that more noise and variability during learning improves transfer over the longer term.

Speakers:
Adele Diamond, University of British Columbia, Vancouver, Canada
Daphne Bavelier, University of Rochester and University of Geneva, Geneva, Switzerland
Torkel Klingberg, Karolinska Institute, Stockholm, Sweden
Silvia A. Bunge, University of California, Berkeley
Amishi P. Jha, University of Miami

Highlights

• Discipline and repeated practice are key to improving executive function (EF).
• Training in EF might reduce achievement gaps.
• Players of action video games show transferable improvements in distractor suppression and attention control.
• Working memory (WM) can be improved, and the improvements are transferrable to new tasks.
• Changes in WM are associated with plasticity in parietal and prefrontal cortex.

We are what we do

Many activities, from mindfulness training to music lessons to martial arts, can promote executive function. Discipline and repetition are the crucial factors, suggested Adele Diamond of the University of British Columbia. "Aristotle realized it: We are what we repeatedly do," she said.

Whether EF improves depends on how an activity is performed, and how frequently. Children with the poorest initial EF generally improve the most regardless of the particularities of the intervention, suggesting that training in this domain is an "excellent candidate for leveling the playing field," she said. Enhancing social, emotional, and physical development in parallel with EF training will be more effective at increasing EF skills than would simply focusing on EF alone. Peer mentoring, individualized attention, active learning, hands-on learning, social engagement, and intrinsic motivation—learning for the sense of mastery rather than to avoid failure—all improve educational outcomes.

Children with the poorest EF improve the most, suggesting EF training is an excellent candidate for leveling the playing field.

Two curricula that incorporate many of these principles improve EF: Tools of the Mind and Montessori. (Diamond argued that Farran's results, which did not find improvement with Tools of the Mind, were compromised by the use of composite outcome measures and should have tracked performance on each individual EF test separately, because some have pronounced ceiling effects that lead to misleading results.) She predicted that musicmaking, play, sports, and dance will be shown to enhance EF because all require persistence, build self-confidence, and promote social belonging. The programs that improve school outcomes will be those that engage children's interests, reduce stress, promote exercise and social acceptance, and provide opportunities to practice EF.

Action video games can promote skills such as attention, visual sensitivity—the ability to distinguish between two similar shades—task switching, approximate mathematics, and general fluid intelligence, said Daphne Bavelier of the University of Rochester and University of Geneva. However, only high-speed action games show these benefits, and not all aspects of cognition change. Action games might offer a unique opportunity to study neuroplasticity and skill acquisition.

Compared to non-games (green bars), players of action video games (blue bars) are faster (lower left chart) and more accurate (lower right chart) at tasks that demand the suppression of distracting information. (Image courtesy of Daphne Bavelier)

Learning is generally quite specific, such that subjects become expert only in the specific task they train to do—chess, solving Rubik's cube puzzles, and so forth—but action games seem to show many transferable benefits that may be sustained by improvements in attentional control. Game-related skills that successfully transfer to other domains include the ability to suppress distracting information and the flexible allocation of attention. "Being able to select signal from noise is critical to being a good learner," said Bavelier. When challenged with demanding cognitive tasks, gamers do not increase activity in frontoparietal networks; much like expert meditators, their brains seem to process information with great efficiency. Possible applications of gaming might include rehabilitation of patients with amblyopia (which results in the loss of binocular vision) or skill training for surgeons and pilots.

Putting working memory to work

Torkel Klingberg of the Karolinska Institute uses parallel genetic, neuroimaging, and behavioral studies to probe factors that predict working memory and techniques that might improve it. Working memory (WM), the ability to keep information readily available for cognitive processes, is deficient in individuals with problems such as ADHD and predicts math and reading ability later on. Allelic variations of two genes, SNAP25 and COMT predict WM, and patterns of brain activity as visualized by neuroimaging during a WM task predict later mathematical ability.

Working memory, deficient in problems like ADHD, predicts math and reading ability.

WM can be improved by training that incorporates repetition and feedback and that is designed to be adaptively difficult so that the challenge steadily increases with performance. WM improvements are associated with increased activity in parietal and prefrontal regions as well as changes in dopamine receptors. In a randomized controlled trial, WM gains transferred to a new task, resulting in significant improvements in attention. In other studies, training of children improved retention of instructions and decreased symptoms of inattention.

Neuroimaging might offer an early opportunity to determine whether an intervention or training protocol promotes learning, suggested Silvia A. Bunge of the University of California, Berkeley. She has investigated the teaching of reasoning skills, which are associated with both academic achievement and general life success. These skills employ a network that includes the dorsolateral prefrontal cortex (DLPFC), ventrolateral prefrontal cortex (VLPFC), parietal cortex, and rostrolateral prefrontal cortex (RLPFC). Teaching reasoning skills to young adults preparing for law school exams led to improvements in LSAT (Law School Admission Test) subscores on portions of the exam that engage logic and reasoning. The 70 hours of training also changed functional connectivity in this network, an effect present even during the resting state when no specific cognition was taking place. Stronger connections developed between the left RLPFC, implicated in relational reasoning, and the posterior parietal cortex, suggesting that the training stimulates neuroplasticity.

Mindfulness practice can improve attentional control, said Amishi P. Jha of the University of Miami. She defined mindfulness as a mental state "characterized by attention to the present moment, without conceptual elaboration or emotional reactivity." Standard practice involves 30 minutes to 45 minutes a day, and engages such cognitive abilities as selecting and maintaining a focus of attention, and being aware of mindwandering. Attention can be thought of like a flashlight, highlighting what is of interest, or as an alerting system to direct attention to potential problems or dangers. In novices, mindfulness training improved the efficiency of directing attention—better control of the flashlight. In another study, mindfulness training improved the ability of subjects to ignore distracting information. These successes in adults suggest mindfulness might improve EF in children. It's "a low cost, low tech, portable technology that might bootstrap other approaches," said Jha.

Speakers:
Eric Pakulak, University of Oregon
Laurie Ford, University of British Columbia, Vancouver, Canada
Frank C. Worrell, University of California, Berkeley
Judy Willis, University of California, Santa Barbara
Susan Magsamen, Johns Hopkins University

Highlights

• Neuroplasticity is a double-edged sword; the flexibility of the brain means that deprived children are subject to cognitive deficits.
• An intervention that pairs parental and child training improves behavior, social skills, and school performance.
• Myths about the brain hinder educational policy.
• Neuroscientists can promote a brain-based view of education.
• Ongoing dialog between educators and neuroscientists is essential.

Poverty and performance

Neuroplasticity, the power of the brain to reshape itself, is generally heralded as a source of potential benefit. Unfortunately, it has a downside: it renders the developing brain sensitive to changing circumstances including situations of deprivation, so that growing up in poverty may result in long-lasting cognitive effects. The good news is that faculties like working memory and attention are malleable during childhood and some, like language ability, can even be modified later in life.

Studies conducted by Eric Pakulak and colleagues of the University of Oregon indicate that while typical three-year-olds modulate brain activity to suppress distracting information, children from low socioeconomic do not show this response. His group studies the relative or synergistic effects of genes and environment on cognition and brain function for attention. They have implemented a training program with Head Start families that teaches stress reduction, consistent discipline, and other skills to parents, while training children in attention, self-regulation, and emotional regulation. The program reduces parent stress, improves parenting behavior, child behavior, child cognition, and child brain function for attention. Gene by environment (intervention) effects are also evident in the data. This suggests that certain variants of genes might not confer vulnerability per se but might confer sensitivity to environment instead, said Pakulak, and training programs could protect against the deleterious effects of that sensitivity.

Gene variants might not confer vulnerability, but sensitivity to environment.

Neighborhood and community are other modifiable aspects of environment that can influence emotional and cognitive development, suggested Laurie Ford of the University of British Columbia. "I want to remind people who work from the neuroscience side to find ways to connect with some of the folks who are doing neighborhood, classroom, and school research," she said.

Frank Worrell of the University of California, Berkeley, said that while researchers have long sought a "magic bullet" to close the achievement gap, many are now coming around to the idea of multiple avenues to address the problem, including modifying environmental conditions, providing cognitive training, and targeting schools and parents. Worrell, who studies gifted students, emphasized the importance of intense effort in success, cautioning that students who lack self-confidence or the sense of self-efficacy may be unwilling to invest sufficient effort to become elite performers. A sense of belonging and powerful intrinsic motivation are important, as well as opportunities and financial resources. Understanding academic achievement should include studying gifted and high-achieving students as well, Worrell noted.

The "achievement gap" refers to lower reading, math, and civics scores among some minorities, but poor children in general score lower, as reflected in the bar second from the right, which charts performance of students eligible for free or reduced-price school lunch. (Image courtesy of Frank Worrell)

Neuroscience in the classroom

Judy Willis of the University of California, Santa Barbara said that she was motivated to change careers to education when she found that many schoolchildren referred to her neurology clinic with putative behavior problems simply seemed overwhelmed and overloaded by tedious schoolwork, she said. They were "acting out or zoning out not because they were bad kids, but because their brains were responding to the stress of boredom and inactivity," she said.

Children's attention in the classroom can be engaged by teaching that taps into curiosity and novelty, essential parts of the brain's attention systems. Dopamine is crucial to attention and learning, and correct predictions drive pleasurable dopamine release. This knowledge can be exploited in a learning curriculum that engages students in predictive tasks. Feedback is also crucial to sustaining attention, and it can be achieved by making tasks incrementally more difficult. "The brain only gets that 'I got it' feeling with greater challenge," she said. "It needs the harder work." Susan Magsamen of Johns Hopkins University described another effort to disseminate research-validated ideas to a wider audience, fostering arts and creative play in "Ultimate block party" events that encourage curiosity and risk-taking in children. Similar ideas inspired the web portal called the Learning Resource Network (L-rn), which will provide practical, evidence-based information about learning for parents and educators.

Despite good intentions and compelling research, there is still a divide between neuroscientists exploring education-related topics and the educators and policymakers who shape classroom agendas. Bridging this gap will require dedicated effort, said Carl E. Wieman of the White House Office of Science and Technology Policy in his keynote address. Research findings do not automatically translate into policy changes. "One problem is that policymakers do not now recognize that cognitive neuroscientists have particular expertise and authority in education," he said.

Policymakers do not now recognize that neuroscientists have particular expertise and authority in education, said Wieman.

In addition, misunderstandings about the brain derail educational strategies. "The lack of expertise leads to a lot of misconceptions about learning that permeate government and the public," he said. Among them: the belief that ability is inborn, that reading comes naturally, that young children can't think abstractly, and particularly, the belief that brains are fixed at birth. These myths feed into misunderstandings like the idea that remedial courses should focus solely on memorization and preparation for real learning.

What scientists can do

A campaign to reconceptualize learning as the result of complex biological processes of the brain could promote neuroscience to a powerful position in educational policy, suggested Wieman. Neuroscientists interested in improving education would do well to reframe education as a domain that requires input from brain science. In his experience developing science, technology, engineering, and math (STEM) curricula for college-aged students, Wieman said that reaching across disciplinary boundaries requires patience and repeated efforts.

The conversation and panel discussion that followed focused on this translational problem in neuroscience of education. "What researchers do far too often is to say, 'We have this great research result, and people out there should do it'," said Wieman, but simply announcing results is not enough. Educators may be unimpressed by what they perceive to be yet another fad. For others, these ideas may be unfamiliar. Educators made up about half the conference audience, but when Willis asked how many had been taught any cognitive science in their educational training, only two raised their hands.

A few concrete ideas came to the fore. Wieman said that National Research Council studies are typically well-regarded in policymaking circles, and could promote evidence-based ideas from neuroscience. John Mighton suggested deriving 10 principles of learning from the presentations given at this conference and using them to craft a curriculum that could be thoroughly tested and disseminated. "There needs to be central body of well-respected evidence that comes in an organized way," he said. Seidenberg agreed that individual efforts are not enough. "There need to be additional structures by which these very different groups can come together."

Goswami cautioned that such a campaign might be premature, and that neuroscience was now in a better position to strengthen the theoretical underpinnings of education rather than to generate specific practical recommendations. "Neuroscience is only one contributor to the enterprise of effective education," she said, "Knowing how to have a curriculum is a different ballgame—you need input from other social scientists." Encouraging teachers' colleges to incorporate neuroscience-based ideas into their curricula is one possibility.

This is a "very exciting, rapidly developing field with enormous potential applications," said Wieman. The conference as a whole fostered the sense that the field of cognitive neuroscience is only beginning to shape educational practice. The scientific presentations, taken together, identified a few key themes such as the nature and duration of cognitive plasticity in children, the best ways to maintain and induce plasticity, and the transferability of skills such as working memory and attention. These questions will drive future research—and further efforts to translate research into a meaningful policy agenda.

Session 1

Can training in executive function (EF) help disadvantaged children prepare for school?

How much can EF be cultivated, and what are the best ways to develop it?

Why do children with one developmental disability tend to experience aspects of others as well?

Why didn't the Tools of Mind curriculum improve EF and academic achievement in a clinical trial?

Can neuroimaging illuminate the gap between genetic risk for a developmental disorder and the phenotype of the disorder?

Session 2

Why is the academic "achievement gap" between African-American children and others so persistent?

Could providing more context reduce the extra cognitive work that African-American children perform when learning to read?

Could early training with rhymes and poems help correct the deficits that lead to dyslexia?

Session 3

Does acuity in the approximate number system underlie better skill in formal mathematics?

How does the brain learn to process quantities and numerals the same way?

Is the enormous variability in math skills among children natural, the effect of schooling, or both?

To what degree is learning transfer—the application of a familiar set of skills to a new problem—possible?

Session 4

What is the best way to build video games and use technology to facilitate transfer of skills from one domain to another?

What cognitive functions besides working memory are plastic and trainable?

Sessions 5 & 6

What is the longterm effect of early enrichment on school performance?

How long do abilities like working memory and attention remain plastic?

What is the best way to bring the expertise of cognitive neuroscientists to the educational policy field?