How People Learn is the product of a 2-year project during which 16 individuals, as a committee, evaluated new developments in the science of learning. We had the good fortune of working with a number of people outside the committee who shared our enthusiasm for this project and we are indebted to the intellectual insights and support that they provided in a number of ways.

A good deal of the excitement that surrounded the project was due to people's seeing the relevance of basic science to education. In light of that connection, the committee held a workshop in fall 1996—"The Science of Science Learning"—to broaden its understanding of the influences that cognitive science has had on science and mathematics learning and teaching. We benefited greatly from the stimulating papers and discussions that grew out of that meeting, as have others who since have used the model of the workshop. We extend our thanks especially to the following people who presented papers and led discussions during the workshop: Susan Carey, Department of Psychology, New York University; Orville L. Chapman, Department of Chemistry, University of California, Los Angeles; Kevin Dunbar, Psychology Department, McGill University; Jill H. Larkin, Department of Psychology, Carnegie-Mellon University; Kevin Miller, Beckman Institute, University of Illinois; Edward F. Redish, Department of Physics and Astronomy, University of Maryland; Leona Schauble, Department of Educational Psychology, University of Wisconsin, Madison; Lee S. Shulman, Stanford University School of Education; Herbert A. Simon, Department of Psychology, Carnegie-Mellon University; and Philip Uri Treisman, Dana Center for Mathematics and Science Education, University of Texas, Austin.

Individually and collectively, members of the committee had discussions with experts on many issues and topics. We wish to acknowledge especially the people who offered suggestions for ways to expand or otherwise improve our collective thinking. In particular, we appreciate the assistance that Ann Rosebery and Beth Warren, both at TERC, Cambridge, MA,

provided on issues of science learning and teaching. Catherine A. Brown, Associate Dean for Research and Development at Indiana University's School of Education, was helpful in sharpening the discussion on mathematics learning and teaching. We also had helpful assistance from Robbie Case, Institute of Child Study, University of Toronto, on issues of children's thinking and from Robert Siegler, Department of Psychology, Carnegie-Mellon University, on children's strategies for learning. Our work on teacher learning and professional development benefited from suggestions provided by Allan Feldman, School of Education, University of Massachusetts.

Although the project was an intellectually exciting undertaking for the committee, we were also mindful of the important role of our sponsor. The Office of Educational Research and Improvement of the U.S. Department of Education established the committee's charge to review the nation's investment in research and the challenge of determining how that investment can pay high returns. We thank Joseph Conaty, Judith Segal, and C. Kent McGuire for the support they provided to this committee in their individual and official capacities.

This report has been reviewed by individuals chosen for their diverse perspectives and technical expertise, in accordance with procedures approved by the Report Review Committee of the National Research Council (NRC). The purpose of this independent review is to provide candid and critical comments that will assist the authors and the NRC in making the published report as sound as possible and to ensure that the report meets institutional standards for objectivity, evidence, and responsiveness to the study charge. The content of the review comments and draft manuscript remain confidential to protect the integrity of the deliberative process.

We wish to thank the following individuals, who are neither officials nor employees of the NRC, for their participation in the review of this report: Kenji Hakuta, School of Education, Stanford University; Donald Kennedy, Institute for International Studies, Stanford University; R. Duncan Luce, Institute for Mathematical Behavioral Science, University of California, Irvine; Michael Martinez, Department of Education, University of California, Irvine; Kevin Miller, Department of Psychology, University of Illinois; Michael I. Posner, Department of Psychology, University of Oregon; Leona Schauble, School of Education, University of Wisconsin, Madison; Herbert A. Simon, Department of Psychology, Carnegie-Mellon University; Patrick Suppes, Professor of Philosophy (emeritus), Stanford University; and Richard F. Thompson, Neurosciences Program, University of Southern California. Although these individuals provided many constructive comments and suggestions, responsibility for the final content of this report rests solely with the authoring committee and the NRC.

Finally, there are several NRC staff and others who made significant contributions to our work. Alexandra Wigdor, director of the Division of

Education, Labor, and Human Performance, of the NRC's Commission on Behavioral and Social Sciences and Education (CBASSE), provided the initial impetus for this project and nurtured it in many different ways that were indispensable to its completion. Eugenia Grohman, associate director for reports of CBASSE, patiently worked with us through several drafts of the report and significantly improved the text. Key support in facilitating our work came from Jane Phillips, senior project assistant in CBASSE, with assistance from Neale Baxter; Susan M. Coke, division administrative associate; Faapio Poe, administrative assistant, Vanderbilt University; and Carol Cannon, administrative assistant, University of California, Berkeley. All of these "behind the scenes" people played critical roles, and to each of them we are very grateful.

The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars engaged in scientific and engineering research, dedicated to the furtherance of science and technology and to their use for the general welfare. Upon the authority of the charter granted to it by the Congress in 1863, the Academy has a mandate that requires it to advise the federal government on scientific and technical matters. Dr. Bruce M. Alberts is president of the National Academy of Sciences.

The National Academy of Engineering was established in 1964, under the charter of the National Academy of Sciences, as a parallel organization of outstanding engineers. It is autonomous in its administration and in the selection of its members, sharing with the National Academy of Sciences the responsibility for advising the federal government. The National Academy of Engineering also sponsors engineering programs aimed at meeting national needs, encourages education and research, and recognizes the superior achievements of engineers. Dr. William A. Wulf is president of the National Academy of Engineering.

The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the services of eminent members of appropriate professions in the examination of policy matters pertaining to the health of the public. The Institute acts under the responsibility given to the National Academy of Sciences by its congressional charter to be an adviser to the federal government and, upon its own initiative, to identify issues of medical care, research, and education. Dr. Kenneth I. Shine is president of the Institute of Medicine.

The National Research Council was organized by the National Academy of Sciences in 1916 to associate the broad community of science and technology with the Academy's purposes of furthering knowledge and advising the federal government. Functioning in accordance with general policies determined by the Academy, the Council has become the principal operating agency of both the National Academy of Sciences and the National Academy of Engineering in providing services to the government, the public, and the scientific and engineering communities. The Council is administered jointly by both Academies and the Institute of Medicine. Dr. Bruce M. Alberts and Dr. William A. Wulf are chairman and vice chairman, respectively, of the National Research Council.

Learning is a basic, adaptive function of humans. More than any other species, people are designed to be flexible learners and active agents in acquiring knowledge and skills. Much of what people learn occurs without formal instruction, but highly systematic and organized information systems—reading, mathematics, the sciences, literature, and the history of a society—require formal training, usually in schools. Over time, science, mathematics, and history have posed new problems for learning because of their growing volume and increasing complexity. The value of the knowledge taught in school also began to be examined for its applicability to situations outside school.

Science now offers new conceptions of the learning process and the development of competent performance. Recent research provides a deep understanding of complex reasoning and performance on problem-solving tasks and how skill and understanding in key subjects are acquired. This book presents a contemporary account of principles of learning, and this summary provides an overview of the new science of learning.

In the last 30 years, research has generated new conceptions of learning in five areas. As a result of the accumulation of new kinds of information about human learning, views of how effective learning proceeds have shifted from the benefits of diligent drill and practice to focus on students' understanding and application of knowledge.

2.  

3.  

4.  

5.  

By definition, experts have developed particular ways to think and reason effectively. Understanding expertise is important because it provides insights into the nature of thinking and problem solving. It is not simply general abilities, such as memory or intelligence, nor the use of general strategies that differentiate experts from novices. Instead, experts have acquired extensive knowledge that affects what they notice and how they

organize, represent, and interpret information in their environments. This, in turn, affects their abilities to remember, reason, and solve problems.

Key scientific findings have come from studies of people who have developed expertise in areas such as chess, physics, mathematics, electronics, and history. The examples are important not because all school children are expected to become experts in these or any other areas, but because the study of expertise shows what the results of successful learning look like.

Key conclusions:

• Experts notice features and meaningful patterns of information that are not noticed by novices.
• Experts have acquired a great deal of content knowledge that is organized, and their organization of information reflects a deep understanding of the subject matter.
• Experts' knowledge cannot be reduced to sets of isolated facts or propositions but, instead, reflects contexts of applicability, i.e., it is ''conditionalized."
• Experts are able to retrieve important aspects of their knowledge with little attentional effort.
• Though experts know their disciplines thoroughly, this does not guarantee that they are able to instruct others about the topic.
• Experts have varying levels of flexibility in their approaches to new situations.

Another aspect of effective learning is its durability—does the learning have long-term impact in the ways it influences other kinds of learning or performance? Research studies on the concept of transfer of learning comprise a vast literature that can be synthesized into the new science of learning.

Key conclusions:

• Skills and knowledge must be extended beyond the narrow contexts in which they are initially learned. For example, knowing how to solve a math problem in school may not transfer to solving math problems in other contexts.
• It is essential for a learner to develop a sense of when what has been learned can be used—the conditions of application. Failure to transfer is often due to learners' lack of this type of conditional knowledge.
• Learning must be guided by generalized principles in order to be widely applicable. Knowledge learned at the level of rote memory rarely transfers; transfer most likely occurs when the learner knows and understands

While there are remarkable commonalties across learners of all ages, children differ from adult learners in many ways. Studies of young children offer a window into the development of learning, and they show a dynamic picture of learning as it unfolds over time. A fresh understanding of infant cognition and of how young children build on early learning predispositions also offers ideas on ways to ease their transition into formal school settings.

Key findings:

• Humans have a predisposition to learn in certain domains, and young children actively engage in making sense of their worlds. In particular domains, such as biological and physical causality, number, and language, infants and young children have strong predispositions to learn rapidly and readily. These biases toward learning support and may make early learning possible and pave the way for competence in early schooling.
• Children lack knowledge and experience, but not reasoning ability. Although young children are inexperienced, they reason facilely with the knowledge they have.
• Precocious knowledge may jump-start the learning process, but because of limited experience and undeveloped systems for logical thinking, children's knowledge contains misconceptions. Misinformation can impede school learning, so teachers need to be aware of the ways in which children's background knowledge influences what they understand. Such awareness on the part of teachers will help them anticipate children's confusion and recognize why the children have difficulties grasping new ideas.

• Strategies for learning are important. Children can learn practically anything by sheer will and effort, but when required to learn about non-privileged domains, they need to develop strategies of intentional learning.
• Children need to understand what it means to learn, who they are as learners, and how to go about planning, monitoring, and revising, to reflect upon their learning and that of others, and to learn to determine for themselves if they understand. These skills of metacognition provide strategic competencies for learning.
• Children are both problem solvers and problem generators. They attempt to solve problems presented to them, and they seek novel challenges. They refine and improve their problem-solving strategies in the face of failure and often build on prior successes. They persist because success and understanding are motivating in their own right.
• Adults help children make connections between new situations and familiar ones. Children's curiosity and persistence are supported by adults who direct children's attention, structure experiences, support learning attempts, and regulate the complexity and difficulty levels of information for children.

Children, thus, exhibit capacities that are shaped by environmental experiences and the individuals who care for them. Developmental processes involve interactions between children's early competencies and the environmental supports—strengthening relevant capacities and pruning the early abilities that are less relevant to the child's community. Learning is promoted and regulated by both the biology and ecology of the child; learning produces development.

Advances in neuroscience are confirming many theoretical hypotheses, including the important role of early experience in development. What is new, and therefore important for a new science of learning, is the convergence of evidence from a number of scientific fields. As developmental psychology, cognitive psychology, and neuroscience, to name but three, have contributed vast numbers of research studies, details about learning and development have converged to form a more complete picture of how intellectual development occurs. Clarification of some of the mechanisms of learning by neuroscience advanced with the advent of non-invasive imaging technologies, such as positron emission tomography (PET) and functional magnetic resonance imaging (FMRI). These technologies enabled researchers to observe directly functions of human learning.

The key finding is the importance of experience in building the structure of the mind by modifying the structures of the brain: development is not solely the unfolding of preprogrammed patterns. Some of the rules that govern learning are now known. One of the simplest rules is that practice increases learning and there is a corresponding relationship between the amount of experience in a complex environment and the amount of structural change in the brain.

Key conclusions:

• Learning changes the physical structure of the brain.
• Structural changes alter the functional organization of the brain; in other words, learning organizes and reorganizes the brain.
• Different parts of the brain may be ready to learn at different times.

Theoretical physics does not prescribe the design of a bridge, but surely it constrains the design of successful ones. Similarly, learning theory provides no simple recipe for designing effective learning environments, but it constrains the design of learning environments—questions that suggest the value of rethinking what is taught, how it is taught, and how it is assessed.

A fundamental tenet of modern learning theory is that different kinds of learning goals require different approaches to instruction; new goals for education require changes in opportunities to learn. The design of learning environments is linked to issues that are especially important in the processes of learning, transfer, and competent performance. Those processes, in turn, are affected by the degree to which learning environments are student centered, knowledge centered, assessment centered, and community centered.

Key conclusions:

• Learner-centered environments Effective instruction begins with what learners bring to the setting; this includes cultural practices and beliefs, as well as knowledge of academic content. A focus on the degree to which environments are learner centered is consistent with the evidence showing that learners use their current knowledge to construct new knowledge and that what they know and believe at the moment affects how they interpret new information. Sometimes learners' current knowledge supports new learning; sometimes it hampers learning.

  People may have acquired knowledge yet fail to activate it in a particular setting. Learner-centered environments attempt to help students make connections between their previous knowledge and their current academic tasks. Parents are especially good at helping their children make connections.

Expertise of any kind involves more than a set of general problem-solving skills; it also requires well-organized knowledge of concepts and inquiry procedures. Various disciplines are organized differently and have

their own methods of inquiry. For example, the evidence needed to support a set of historical claims is different from the evidence needed to prove a mathematical conjecture, and both of these differ from the evidence needed to test a scientific theory.

Key conclusions:

• Effective teachers need "pedagogical content knowledge"—knowledge about how to teach in particular disciplines, which is different from knowledge of general teaching methods.
• Expert teachers know the structure of their disciplines and this provides them with cognitive roadmaps that guide the assignments they give students, the assessments they use to gauge student progress, and the questions they ask in the give and take of classroom life.

In short, teachers' knowledge of the discipline and their knowledge of pedagogy interact. But knowledge of the discipline structure does not in itself guide a teacher. Expert teachers are sensitive to those aspects of the discipline that are especially hard and easy for new students to master. An emphasis on interactions between disciplinary knowledge and pedagogical knowledge directly contradicts a common misconception about what teachers need to know in order to design effective learning environments for their students. The misconception is that teaching consists only of a set of general methods, that a good teacher can teach any subject, and that content knowledge alone is sufficient.

Teacher learning is relatively new as a research topic, so there is limited information about it. Nevertheless, the research that exists, generally in the form of rich case studies, provides important information about what kinds of learning opportunities teachers need in order to change their practices.

Key findings:

• Opportunities for teachers to continue their learning fall short when viewed from the perspective of being learner, knowledge, assessment, and community centered. Preservice programs often fail to provide the kinds of learning experiences that lead to learning for understanding or teaching for understanding.
• Successful learning for teachers requires a continuum of coordinated efforts that range from preservice education to early mentored teaching to opportunities for lifelong development as professionals. Creating such opportunities represents a major challenge.

A number of the features of new technologies are consistent with the principles of a new science of learning.

Key conclusions:

• Because many new technologies are interactive, it is now easier to create environments in which students can learn by doing, receive feedback, and continually refine their understanding and build new knowledge.
• Technologies can help people visualize difficult-to-understand concepts, such as differentiating heat from temperature. Students are able to work with visualization and modeling software similar to the tools used in nonschool environments to increase their conceptual understanding and the likelihood of transfer from school to nonschool settings.
• New technologies provide access to a vast array of information, including digital libraries, real-world data for analysis, and connections to other people who provide information, feedback, and inspiration, all of which can enhance the learning of teachers and administrators as well as students.

There are many ways that technology can be used to help create such environments, both for teachers and for the students whom they teach. However, many issues arise in considering how to educate teachers to use new technologies effectively. What do they need to know about learning processes? About the technology? What kinds of training are most effective for helping teachers use high-quality instructional programs? What is the best way to use technology to facilitate teacher learning? Good educational software and teacher-support tools, developed with full understanding of principles of learning, have not yet become the norm.

It will take time and effort to communicate the new approaches to learning and teaching throughout the very decentralized U.S. education system. We suggest a number of ways to begin the process through a research agenda that follows from our conclusions. The research will have greatest potential for impact in education if it is implemented as a program of research, making educational research an integrative science.

• The committee recommends a commitment to basic research programs in cognition, learning, and teaching.

Our report has shown the payoff from investing in research on such topics as the foundational role of learners' prior knowledge in acquiring new information; plasticity and adaptability of learning; the importance of social and cultural contexts in learning; understanding the conditions of

transfer of learning; how the organizational structure of a discipline affects learning; how time, familiarity, and exploration impact fluency in learning; and many other topics. While these areas have produced an impressive body of research findings, the research needs to be continued. The framework has been constructed from the earlier research; details now need to be provided in order to advance the science of learning by refining the principles.

• The committee recommends establishing new research programs in emerging areas, including technology, neurocognition, and sociocultural factors that mediate learning. Research is needed on the interrelations between learning and learning environments and between teaching and learning. This research will build on current findings in areas such as how children learn to apply their competencies as they encounter new information; how early competencies relate to later school learning; the conditions and experiences that support knowledge scaffolding; and how representational systems are challenged by new tools of technology, such as visual cognition and other types of symbolic thinking.
• The committee recommends new assessment research to focus on improving and implementing formative assessment. Teachers need a variety of supports and learning opportunities for making their classrooms assessment centered in ways that support learning. Research questions that remain to be addressed include: How does a teacher use assessment? What skills do teachers need in order to be able to use formative assessments in ways that will improve their teaching? What kinds of supports do teachers need for learning and adopting innovative assessment processes?

The committee held a workshop on children's cognitive development and the ways in which cognitive science research has influenced science instruction in recent years. The workshop explored ways in which new research findings can facilitate new directions in areas of science and mathematics learning.

Key questions:

• How does the field "scale up" successful demonstrations of research-based curricula so that they can be implemented in many diverse settings under the guidance of many different kinds of teachers?
• Which factors influence the conversion of research knowledge into effective instructional methods in real settings?

• Do strategies that work for science education also work to improve instruction in other subject areas?
• How can preschool children be assisted in developing representational structures so that there are bridges, rather than gaps, between early and later school learning?
• How can collaborative learning environments be organized in ways that counteract societal stereotypes and tap diversity as a positive resource for learning?
• Which kinds of assessments can effectively measure new kinds of science learning?
• How do the features of a constructivist curriculum interact with other social factors in classrooms?
• What is the impact of new technologies on school performance?

The research areas relevant to the science of learning are demonstratively broad, including cognitive development, cognitive science, developmental psychology, neuroscience, anthropology, social psychology, sociology, cross-cultural research, research on learning in subject areas such as science, mathematics, history, and research on effective teaching, pedagogy, and the design of learning environments. New technologies are needed for assessing learning in ways that track the growth of learning, not just the cumulation of facts. Developing effective research methodologies is particularly important for research from this diverse array of disciplines.

• The committee recommends that government agencies and research foundations develop initiatives and mechanisms of support specifically aimed at strengthening the methodological underpinnings of learning sciences. Such mechanisms should include cross-field collaborations, internships, visiting scholar programs, training junior scholars in interdisciplinary approaches, and other procedures to foster collaborations for learning and developing new methodologies that can lead to more rigorous investigations in the science of learning.
• The committee recommends research aimed at developing and standardizing new measures and methods. Studies should be conducted and validated with diverse populations. New statistical techniques should be developed for analyzing the complex systems of learning. New qualitative measurement techniques need to be developed.
• The committee recommends new research that is focused on ways to integrate qualitative and quantitative methods across the learning sciences.

This book emphasizes the breadth of knowledge areas that affect learners and the significant advances that have been the direct result of collaborative research efforts across disciplines. That kind of collaboration is critical to further development of the learning sciences.

• The committee recommends that government agencies and research foundations explicitly support a wide variety of interdisciplinary collaborations in the learning sciences. Such work should include teachers.

  The field of learning research needs to become more integrated in focus and draw together relevant fields for interdisciplinary collaborations. To this end, mechanisms are needed to prepare a new generation of learning scientists by supporting interdisciplinary training for students and scientists to work together. It is important to expand the research scope so that basic researchers and educational researchers can work together on basic and applied issues and to facilitate ways for teachers and researchers to work together. Fields such as neuroscience and cognitive science have made important advances through their joint efforts, but researchers had to learn the methodologies and techniques of each discipline before new research studies could be conducted. Efforts are now needed to direct training programs in order to foster such interdisciplinary learning.

• The committee recommends establishing national databases to encourage collaboration.

  To capitalize on the new developments in information systems, research scientists of varying disciplines should be linked together, and teachers should be included in these virtual dialogues. In addition to electronic linkages through websites, scientists should begin to share databases with one another and work with national databases that they can access electronically.

  Databases that link physics researchers with classroom physics educators, for example, have the potential to bring the two sectors closer to the core issues of the field. Basic researchers often have a poor understanding of why learners fail to grasp basic concepts of the field; teachers often fail to see relationships of core concepts that, if better understood from the standpoint of theory, could facilitate their teaching. National databases can foster interdisciplinary collaboration and uses of cross-disciplinary data, promote broader exploration of testable questions across datasets, increase the quality of data by maintaining accurate and uniform records, and promote cost-effectiveness through the sharing of research data. Furthermore, national databases that are built from representative samples of the changing school population have the potential of broadening the scope and power of research findings.

Because many computer-based technologies are relatively new to classrooms, basic premises about learning with these tools need to be examined with respect to the principles of learning.

• The committee recommends extensive evaluation research be conducted through both small-scale studies and large-scale evaluations to determine the goals, assumptions, and uses of technologies in classrooms and the match or mismatch of these uses with the principles of learning and the transfer of learning.

Much of what constitutes the typical approach to formal teacher professional development is antithetical to what promotes teacher learning.

• The committee recommends research to explain how people learn to be effective teachers.

Research studies are needed to determine the efficacy of various types of professional development activities, including preservice and inservice seminars, workshops, and summer institutes. Studies should include professional activities that are extended over time and across broad teacher learning communities in order to identify the processes and mechanisms that contribute to the development of teachers' learning communities.