4

Proposing a Research and Development Agenda

     In designing a research and development agenda, the committee considered the mechanisms through which research influences practice, in light of the feedback received at the two events organized around How People Learn. Recall that Figure 1.1 from Chapter 1 depicts the committee's view of the paths through which research influences practice.

     Several aspects of the figure are worth noting again here. First, the influence of research on the four mediating arenas--education materials, pre-service and in-service teacher and administrator education programs, public policy, and public opinion and the media--has typically been weak for the variety of reasons discussed earlier. Without clear communication of a research-based theory of learning and teaching, the operational theories held by the various stakeholders are not aligned. Teachers, administrators, and parents frequently encounter conflicting ideas about the nature of learning and its implications for effective teaching.

     Second, with the exception of the relatively small set of cases in which teachers and researchers work together on design experiments, the arrows between research and practice in Figure 1.1 are one-way. This reflects the fact that practitioners typically have few opportunities to shape the research agenda and contribute to an emerging knowledge base of learning and teaching. The task of bridging research and practice requires an agenda that allows for a flow of information, ideas, and research questions in both directions. It requires an agenda that consolidates the knowledge base and strengthens the links between that knowledge base and each of the components that together influence practice.

     The potential benefits of bridging theory and practice are noted by Donald Stokes in his recent work, Pasteur's Quadrant (1997). Stokes observed that many of the advances in science are intimately connected to the search for solutions to practical problems. Pasteur appears in the book's title because his work contributed so clearly to scientific understanding while simultaneously focusing on practical problems. Such research is "use-inspired." As in Pasteur's case, when executed as part of a systematic and strategic program of inquiry, it can support new understandings at the most fundamental and basic scientific level.

     A central theme of Stokes's argument is that the typical linear conceptualization of research as a sequence from basic to applied is an inaccurate characterization of much research, and it is highly limiting for the envisioning of a research agenda. He proposes instead a quadrant in two-dimensional space in which considerations of use and the quest for fundamental understanding define the horizontal and vertical axes respectively. The quadrant allows for the possibility that research can be high in both basic and applied values.

     The committee is sympathetic to this perspective. We envision the need for a comprehensive program of use-driven strategic research and development focused on issues of improving classroom learning and teaching. The facts that schools and classrooms are the focus and that enhanced practice and learning are the desired goals render the program of research no less important with respect to advancing the theoretical base for how people learn. Indeed, many of the advances described in How People Learn are the product of use-inspired research and development focused on solving problems of classroom practice.

     It is worth noting that a wide array of quantitative and qualitative methods drawn from the behavioral and social sciences are employed in education research. The methods often vary with the nature of the learning and teaching problem studied and the level of detail at which issues are pursued. Given the complexity of educational issues in real-world contexts in which variables are often difficult to control, the types of "use-inspired" research that we envision will necessarily demand a variety of methods. These will range from controlled designs to case studies, with analytic methods for deriving conclusions and inferences including both quantitative and qualitative procedures of substantial rigor. To build an effective bridge between research and practice, such a multiplicity of methods is not only reasonable, it is essential. No single research method can suffice.

     Adopting the perspective of use-inspired, strategic research and development focused on issues of learning and teaching is a powerful way to organize and justify the specific project areas we describe. In light of the many comments of workshop participants, the committee identified five overarching themes that guided our understanding of the change that is required to bridge research and practice more effectively. Three of these themes point to the consolidation of knowledge that would help link research and practice:

     1. Elaborate the messages in How People Learn at a level of detail that makes them usable to educators and policy makers. Workshop participants were enthusiastic about the report and its implications for classroom teaching. They were virtually unanimous, however, in the view that the findings and their implications need to be substantially elaborated and incorporated into curricula, instructional tools, and assessment tools before their impact will be felt in the classroom. It is not enough to know, for example, that subject-matter information must be tied to related concepts if deep understanding and transfer of learning are the goals. Teachers must recognize which particular concepts are most relevant for the subject matter that they teach. And they need curriculum materials that support the effort to link information with concepts. Similarly, policy makers need to know quite specifically how the principles in How People Learn relate to state standards. In this sense, the development aspect of the agenda is critical.

     2. Communicate the messages in How People Learn in the manner that is most effective for each of the audiences that influences educational practice. For teachers to teach differently and administrators and policy makers to support a different model of teaching, they need opportunities to learn about the recommended changes and to understand what they are designed to achieve. Research must be done on effective methods of communicating these ideas to teachers, administrators, and policy makers, each of whom have different information needs and different ways of learning. Similarly, teachers, administrators, and policy makers all emphasized that the public's beliefs regarding education influence how they do their jobs. They recommended research aimed at effectively communicating key ideas from How People Learn to the public.

     3. Use the principles in How People Learn as a lens through which to evaluate existing education practices and policies. How People Learn emphasizes that many existing school practices and policies are inconsistent with what is known about learning. But there are also havens of exemplary educational practice, and the report points to some of these as well. The education landscape is dotted with reform efforts and with institutes and centers that produce new ideas and new teaching materials. Educators, administrators, and policy makers are eager for help in sorting through what already exists. They want to know which of these current practices, training programs, and policies are in alignment with the principles in How People Learn, and which are in clear violation of them.

     Moreover, educators emphasized that new ideas are introduced to schools one after another, and teachers become weary and skeptical that any new reform effort will be better than the last. Zealous efforts to promote the newest idea often overlook existing practices that are successful. An effort to identify such practices will build support from those who have long been engaged in teaching for understanding.

     Together, these three themes suggest that an effective bridge between research and practice will require a consolidated knowledge base on learning and teaching that builds, or is cumulative, over time. Elaborating on the committee's conceptualization in Figure 1.1, this knowledge base appears at the center of Figure 4.1. Fed by research, it organizes, synthesizes, interprets, and communicates research findings in a manner that allows easy access and effective learning for those in each of the mediating arenas. Attending to the communication and information links between the knowledge base and each of the components of the model simultaneously enhances the prospect for the alignment of research ideas and practice.

     Two additional themes that emerged from the discussions focus on how research should be conducted to strengthen its link to practice:

     4. Conduct research in teams that combine the expertise of researchers and the wisdom of practitioners. Much of the work that is needed to bridge research and practice focuses on the education and professional development of teachers, the curriculum, instruction and assessment tools that support their teaching, and the policies that define the environment in which teaching takes place. These are areas about which practitioners have a great deal of knowledge and experience. Workshop participants emphasized the need to have educators partnered with researchers in undertaking these research projects. Such partnerships allow the perspectives and knowledge of teachers to be tapped, bringing an awareness to the research of the needs and dynamics of a classroom environment. Since such partnerships are novel to many researchers, exemplary cases and guiding principles will need to be developed to make more likely the successful planning and conduct of research team partnerships.

     5. Extend the frontier of learning research by expanding the study of classroom practice. Researchers and practitioners who participated in the workshops recommended expansion of research efforts that begin by observing the learning that takes place in the classroom. This research, as the earlier discussion of the Stokes work suggests, may advance understanding of the science of learning in important and useful ways.

     Taken together, these latter two suggestions imply that the links between research and practice should routinely flow in both directions. The insights of researchers help shape the practitioner's understanding, and the insights of practitioners help shape the research agenda and the insights of researchers. Moreover, the link between each of the arenas and the knowledge base flows in both directions. Efforts to align teaching materials, teacher education, administration, public policy, and public opinion with the knowledge base are part of an ongoing, iterative research effort in which the implementation of new ideas, teaching techniques, or forms of communication are themselves the subject of study.

     The agenda that follows proposes research and development that can help consolidate the knowledge base (which appears at the center of Figure 2.1) and can build the two-way links between the knowledge base and each of the arenas that influences practice. But that knowledge base is also fed by research on learning more generally and on classroom practice. The committee also suggests additional research that would strengthen the understanding of learning in areas that go beyond How People Learn.

     Finally, since communication and access to knowledge are key to alignment, the committee proposes a new effort to use interactive technologies to facilitate communication of the variety of findings that would emerge from these research and development projects.

     In many of the proposed areas for research and development, work is already under way. Inclusion in the agenda is not meant to overlook the contributions of research already done or in progress. Rather, we are inclusive in order to suggest that research findings need to be synthesized and integrated into the knowledge base and their implications tested through ongoing, iterative research.

     The goal of the recommended research and development in this area is to build on and elaborate findings in How People Learn so that they are "applications ready" and more usable to those responsible for developing curriculum, instructional, and assessment materials. It is designed to achieve three interrelated goals: (a) to identify existing educational materials that are aligned with the principles of learning suggested by the report and to develop and test new materials in areas of need, (b) to advance the knowledge base by significantly extending the work described in How People Learn to additional areas of curriculum, instructional techniques, and assessments that are in need of detailed analysis, and (c) to communicate the messages in How People Learn in a manner appropriate to developers of educational materials and teachers by using a variety of technologies (e.g., texts, electronic databases, interactive web sites). The research we recommend is described in this section in seven project areas.

     1. Review a sample of current curricula, instructional techniques, and assessments for alignment with principles discussed in How People Learn. The committee recommends that teams of discipline-specific experts, researchers in pedagogy and cognitive science, and teachers review a sample of widely used curricula, as well as curricula that have a reputation for teaching for understanding. The envisioned research would involve two stages; these might be conducted together in a project, or as sequential projects.

     Stage 1: These curricula and their companion instructional techniques and assessments should be evaluated with careful attention paid to alignment with the principles of learning outlined in How People Learn. The review might include consideration of the extent to which the curriculum emphasizes depth over breadth of coverage; the effectiveness of the opportunities provided to grasp key concepts related to the subject matter; the extent to which the curriculum provides opportunities to explore preconceptions about the subject matter; the adequacy of the factual knowledge base provided by the curriculum; the extent to which formative assessment procedures are built into the curriculum; and the extent to which accompanying summative assessment procedures measure understanding and ability to transfer rather than memory of fact.

     The features that support learning should be highlighted and explained, as should the features that are in conflict. The report from this research should accomplish two goals. First, it should identify examples of curriculum components, instructional techniques, and assessment tools that incorporate the principles of learning. Second, the explication of features that support or conflict with the principles of learning should be provided in sufficient detail and in a format that allows the report to serve as a learning device for those in the education field who choose and use teaching and assessment tools. As such, it could serve as a reference document when new curricula and assessments are being considered.

     Stage 2: The curricula that are considered promising should be evaluated to determine their effectiveness when used in practice. Curricula that are highly rated on paper may be very difficult for teachers to work with, or in the light of classroom practice may fail to achieve the level of understanding for which they are designed. Measures of student achievement take center stage in this effort. Through the lens of How People Learn, achievement is indicated not only by a command of factual knowledge, but also by a student's conceptual understanding of subject matter and the ability to apply those concepts to future learning of new, related material. If existing assessments do not measure conceptual understanding and knowledge transfer, then this stage will require development and testing of such measures. In addition to achievement scores, feedback from teachers and curriculum directors who use the materials would provide additional input for stage 2.

     Ideally, the review of curricula would take place at several levels: at the level of curriculum units, which may span several weeks of instructional time; at the level of semester-long and year-long sequences of units; and at the level of multiple grades, so that students have chances to progressively deepen their understanding over a number of years.

     The curricula reviewed should not be limited to those that are print based. As a subset of this effort, the committee recommends a review of curricula that are multimedia. The number of computers in schools is expanding rapidly. For schools to use that equipment to support learning, they must be able to identify the computer-based programs that can enhance classroom teaching or class assignments. The committee recommends that research be done to:

     a. Identify technology programs or computer-based curricula that are aligned with the principles of learning for understanding. The programs identified should go beyond those that are add-ons of factual information or that simply provide information in an entertaining fashion. The investigation should explore how the programs can be used as a tool to support knowledge-building in the unit being studied, and how they can further enhance the development of understanding of key concepts in the unit. The study should also explore the adequacy of opportunities for learning about the programs and for ongoing support in using the programs in a classroom setting;

     b. Evaluate the aligned programs as teaching/learning tools by conducting empirical research on their distinctive contribution to achievement and other desired outcomes.

     c. Investigate computer programs that appear to be effective teaching devices but do not clearly align with the principles of learning. These might suggest productive areas for further study.

     2. In areas in which curriculum development has been weak, design and evaluate new curricula, with companion assessment tools, that teach and measure deep understanding. As an extension of Recommendation 1 above, or in some cases as a substitute, the committee recommends the development and evaluation of new curriculum and assessment materials that reflect the principles of learning outlined in How People Learn. Once again, the committee recommends that the development be done by teams of disciplinary experts, cognitive scientists, curriculum developers, and expert teachers. Ideally, research in this category will begin with existing curricula and modify them to better reflect key principles of learning. In some cases, however, exemplary curricula for particular kinds of subject matter may not exist, so the teams will need to create them. This research and development might be coordinated with the ongoing efforts of the National Science Foundation to ensure complementary rather than duplicative efforts.

     The curricula should be designed to support learning for understanding. They will presumably emphasize depth over breadth. The designs should engage students' initial understanding, promote construction of a foundation of factual knowledge in the context of a general conceptual framework, and encourage the development of metacognitive skills.

     Companion teacher materials for a curriculum should include a "meta-guide" that explains its links to principles of learning, reflects pedagogical content knowledge concerning the curriculum, and promotes flexible use of the curriculum by teachers. The guide should include discussion of expected prior knowledge (including typical preconceptions), expected competencies required of students, and ways to carry out formative assessments as learning proceeds. Potentially excellent curricula can fail because teachers are not given adequate support to use them. Although instructional guides cannot replace teacher training efforts, the meta-guide should be both comprehensive and user-friendly to supplement those efforts. Finally, both formative and summative tests of learning and transfer should be proposed as well.

     Once developed, the committee recommends field-testing of the curricula in order to amass data on student learning and teacher satisfaction, identifying areas for improvement. Clearly, it is easier to field-test short units rather than longer ones. Ideally, different research groups that are focusing on similar topics across different age groups (e.g., algebra in elementary, middle, and high school) would work to explore the degree to which each of the parts seems to merge into a coherent whole.

     The committee recommends once again that careful attention be paid to the criteria used to evaluate the learning that is supported by the materials and accompanying pedagogy. Achievement should measure understanding of concepts and ability to transfer learning to new, related areas.

     3. Conduct research on formative assessment. The committee recommends a separate research effort on formative assessment. The importance of making students' thinking visible by providing frequent opportunities for assessment, feedback, and revision, as well as teaching students to engage in self-assessment, is emphasized in How People Learn and in the proposals above. But the knowledge base on how to do this effectively is still weak. To bolster the understanding of formative assessment so that it can more effectively be built into curricula, this research effort should:

     a. Formulate design principles for formative assessments that promote the development of coherent, well-organized knowledge. The goal of these assessments is to tap understanding rather than memory for procedures and facts;

     b. Experiment with approaches to developing in students and teachers a view of formative assessment and self-assessment as an opportunity for providing useful information that allows for growth, rather than as an outcome measure of success or failure;

     c. Explore the potential of new technologies that provide the opportunity to incorporate formative assessment into teaching in an efficient and user-friendly fashion.

     This research effort should consider as well the relationship between formative and summative assessments. If the goal of learning is to achieve deep understanding, then formative assessment should identify problems and progress toward that goal, and summative assessment should measure the level of success at reaching that goal. Clearly they are different stages of the same process and should be closely tied in design and purpose.

     4. Develop and evaluate videotaped model lessons for broadly taught, common curriculum units that appear throughout the K-12 education system. Many lessons and units of study are taught almost universally to students in the United States. Examples include the rain cycle in science, the concept of gravity in physics, the Civil War in history, and Macbeth in English. The committee recommends that a sample of familiar teaching topics be chosen to illustrate teaching methods that are compatible with the findings of How People Learn. The research and development should be undertaken by teams composed of disciplinary experts, pedagogical experts, master teachers, and video specialists. The model lessons or units envisioned by the committee would in all cases:

     a. Illustrate a methodology for drawing out and working with student preconceptions and assessing progress toward understanding (results from project area 5 below could contribute to this endeavor);

     b. Present the conceptual framework for understanding or organizing the new material;

     c. Provide clear opportunities for transfer of knowledge to related areas.

When appropriate, they would also:

     d. Provide instruction on the use of meta-cognitive skills;

     e. Include examples of group processes in the development of understanding, illustrating the nature (and potential advantages) of capitalizing on shared expertise in the classroom.

     The model units would be prefaced and heavily annotated to guide the viewer's understanding. Annotations would include both subject content and pedagogical technique. Companion assessment tools should be developed that measure understanding of the core concepts taught in the lessons.

The committee recommends multiple models of teaching the same unit in different school contexts. These could serve several purposes. First, the goal of the videotaped models is to illustrate effective approaches to teaching more generally, not just of teaching a particular unit. This learning is more likely to occur with multiple examples that allow for variation in the delivery of the lesson, holding constant the underlying principles of effective teaching.

     Second, the classroom dynamics and level of preparation of the students can vary significantly from one school to the next. It may be difficult for a teacher to find relevant instruction in a videotape of a class that does not resemble the one in which she or he teaches. Finally, the art of teaching requires flexibility in responding to students' inquiries and reflections. Multiple cases can demonstrate flexibility in response to the particular students being taught while attending to a common body of knowledge.

     Whether providing multiple models does indeed achieve these purposes is itself a research question worth pursuing. Such research should test the effect of each additional model provided on the level of understanding of key learning and teaching concepts, as well as the amount of variation between models that optimizes the flexibility of understanding that viewers achieve.

     Once pilot versions of these lessons are designed, the committee recommends rigorous field-testing, with time built into the research plan for revision and retesting. Video-based materials already developed and in use as part of the National Board for Professional Teaching Standards training and assessment development process should be considered as possible candidate materials for further study as part of this process.

     The committee recommends that the model lessons be organized in widely accessible video and multimedia libraries that could serve multiple purposes:

• The lessons could be used as anchors for discussions of pre-service and in-service teachers and administrators, as they try to understand and master the pedagogy to accompany the new forms of learning described in How People Learn.

• The lessons could be instructive in administrative training programs. School administrators responsible for hiring and evaluating teachers need models of good practice that can inform their evaluations.

• With some modified annotations, the lessons could inform parents about teaching techniques that promote learning for understanding. Changing classroom teaching can be problematic if new methods run counter to parents' perceptions of the learning process. The model lessons could help parents understand the goals of the espoused approach to teaching.

     5. Conduct research on key conceptual frameworks, by discipline, for the units that are commonly taught in K-12 education. A key finding of the research reviewed in How People Learn is that deep understanding--and the transfer of learning that is one of its hallmarks--requires that the subject matter being taught be tied to the key concepts or organizing principles that the discipline uses to understand that subject. The goal of teaching about a given topic is not simply to convey factual information, although that information is a necessary component. The meaning of that information as it relates to basic concepts in the discipline, the related analytical methods that answer the question "How do we know," and the terms of discourse in a disciplinary field are all components in developing competence.

     To illustrate, consider the topic of marine mammals as it might be taught in early elementary school. That unit would be likely to include identification of the various marine mammals, information on the features that distinguish marine mammals from fish, and perhaps more detailed information on the various types and sizes of whales, the relative size of male and female whales, etc. To the marine biologist, this information is the interesting detail in a larger story, which begins with the question: "Why are there mammals in the sea?" A unit organized around that question would engross students in an evolutionary tale in which the adaptation of sea creatures for life on the land takes a twist: land mammals now adapt to life in the sea. The core biological concepts of adaptation and natural selection would be at the center of the tale. Students would come to understand the puzzle that marine mammals posed for scientists: Could sea creatures evolve to mammals that live on land and then evolve again to mammals that return to the sea? They would come to understand the debate in the scientific community and the discovery of supporting evidence. And they would have cause to challenge the widespread misconception that evolution is a unidirectional process.

     The approach of tying information on marine mammals to the concepts, language, and ways of knowing in that branch of science can be used in other areas of science, as well as in other disciplines. But the concepts and organizing principles that provide a framework for particular subject matter are often obvious only to those who are expert in the discipline. The committee recommends that discipline-specific research be conducted in history, mathematics, natural sciences, and social sciences to systematically review units of study that commonly appear in K-12 curricula, specifying the conceptual framework to which the unit should be tied. The results of this effort will allow teachers and curriculum developers to see if a common conceptual basis exists for separate units of study. Making those underlying concepts explicit helps students construct a model for understanding that facilitates transfer.

     The committee further recommends that the work in each discipline be reviewed by a panel of disciplinary experts to identify consensus and contested areas. To the extent that there is a high level of agreement within a discipline about the organizing constructs as they apply to units of classroom study, the outcome of this research will be highly useful to those who design and evaluate curricula and to those who teach.

     6. Identifying and addressing preconceptions by field. The research reviewed in How People Learn makes the case that new learning is built on the foundation of existing knowledge and preconceived notions regarding the subject of study. Learning is enhanced when preconceived understandings are drawn out. When these are accurate, new knowledge can be directly tied to what is already known. And when they are inaccurate, students can be made aware of how their existing conceptions fall short and be provided with more robust alternatives. Teachers and curriculum developers can build learning experiences into curricula that challenge typical misconceptions, and that draw out and work with unpredictable preconceptions. The committee recommends research by discipline and subject area:

     a. To identify common preconceptions that students bring to the classroom at different levels of education;

     b. To identify links that can be made between existing learner understandings and the disciplinary knowledge, when they are compatible;

     c. To identify progressive learning sequences that would allow students to bridge naïve and mature understandings of the subject matter.

     The research would be conducted independently for mathematics, natural sciences, social sciences, and humanities. The research teams should combine disciplinary experts with cognitive scientists, expert teachers, and curriculum developers. The range of topics covered in each disciplinary area should allow for exploration of the key concepts in the field as they arise in commonly covered course topics in the K-12 curriculum.

     In some disciplines (e.g., physics), substantial research has already been done to identify misconceptions. This project should build on those efforts but extend them by developing and testing strategies for working with preconceptions, providing tools and techniques for teachers to work with in the classroom.

     The research, as envisioned, would involve several stages:

• Stage 1 would involve the identification of the subject areas for study and the key concepts that students must comprehend in order to understand each subject area. Assessment tools that allow for a test of comprehension of these concepts, including tests of the degree to which students' understanding supports new learning (transfer), would also be developed at this stage.

• Stage 2 would consist of a review of existing research that explores the preconceptions that students bring to that subject area and an extension of the research into areas that have not been adequately explored.

• Stage 3 would involve the development of learning opportunities and instructional strategies that build on, or challenge, those preconceptions. These might include experiments in physics that produce results contradicting initial understandings, or research tasks in history that show the same event from multiple perspectives, challenging good-guy/bad-guy stereotyping.

• Stage 4 would involve experimental testing of the newly developed learning tools and instructional strategies, with the assessment tools developed in stage 1 used as a measure of comprehension.

     The final products of this research in each disciplinary area would include written reports of research results, as well as descriptions of tested instructional techniques for working with student preconceptions. The findings could be incorporated into videotaped model lessons (project area 4 above) or those used in the pedagogical laboratories proposed in the project area 13.

     7. Develop an interactive communications site that provides information on curricula by field. Feedback from workshop participants suggested a high level of frustration with the task of sorting through and evaluating curricula. A central source of information on curricula and their major features would be highly valued. The committee recommends the development and maintenance of an interactive communications site that provides information about design principles for effective curricula, and relates these principles to particular curricula by subject area. The curriculum review and development recommended above would provide a solid foundation of information for creation of the site.

     Comparing and rating curricula can be a difficult business. A good curriculum will need to balance coverage of information with in-depth exploration of concepts. But there is no magic balancing point. One curriculum may provide more opportunities to explore interesting scientific narratives, whereas another may offer more opportunities for valuable experimentation. But if the difficulty in evaluating curricula means backing away entirely from the effort to compare and evaluate, then the information available to those who must choose among curricula is diminished. Thousands of schools and teachers must then bear a much heavier burden of information collection.

     The committee recommends a comprehensive evaluation process that does not rank-order curricula, but rather evaluates them on an array of relevant features. A sample of such features taken from How People Learn includes the extent to which the curriculum draws out preconceptions; whether it includes embedded assessment (both formative and summative); the extent to which it places information in the relevant conceptual framework; the extent to which curriculum modules can be reconfigured in ways that allow teachers to meet particular goals and needs, and the extent to which it encourages the development of metacognitive skills. Other useful information on the curriculum would include the extent and results of field-testing, the length of time it has been in use, the number of schools or school districts that have adopted it, the opportunities for teacher learning, and the amount and kind of support available to teachers using the curriculum. Information on student response to and interest in the curriculum would be useful as well.

     Evaluating curricula in terms of their relevant features that align with the principles in How People Learn is a massive undertaking. For its ultimate success, such evaluations will need to represent expert judgments coming from different perspectives, including the subject-matter discipline, master teachers, learning and pedagogy experts, and curriculum developers. Users of an interactive communications site that publishes these judgments can then weigh the expertise they consider most useful for guiding their choice of curricula. The site should invite their feedback on experiences with using the curricula that this information led them to select. Ideally, the communications site will make it easy for teachers to access information that is directly relevant to their particular goals and needs.

     Success will also require a growing group of constituencies and experts who can carry forward the principles in How People Learn to evaluating curricula.

     The research and development proposed in this section is designed, once again, to achieve three goals: (a) to look first at existing practice through the lens of How People Learn, (b) to advance understanding in ways that would facilitate alignment of teacher preparation with principles of learning, and (c) to make the findings of this research more widely accessible and easily understood. The recommended research is described in seven project areas.

     8. Review the structure and practices of teacher education for alignment with the principles of learning. For teacher education and professional development programs to be aligned with the principles of learning, they need to prepare teachers to think about the enterprise of teaching as building on the existing knowledge base and preconceptions of their students, to teach skills for drawing out and working with existing understandings, and to continually assess the progress of students toward the goal of deep understanding. The programs need to provide for their students the opportunity to develop a deep understanding themselves of the subject matter they will teach and the ability to facilitate students' transfer of knowledge to related areas. They need to prepare teachers to be aware of and directly teach metacognitive skills. And they need to convey a model of the teacher as learner, who continually develops expertise that is flexible and adaptive.

     These are implications for what schools of education and professional development programs should teach. But the students in those programs will themselves learn more effectively if they are taught according to these principles. How People Learn therefore has implications for how schools of education do their job. Do those schools have program structures and practices that reflect the principles of learning discussed here?

     The committee recommends that evaluation research be conducted to examine current program structures and practices at schools of education through the lens of How People Learn. This effort should not only synthesize what is already known about teacher training programs, but also undertake a new evaluation. The sample of schools should be chosen to reflect the wide range of program formats (which currently include undergraduate and postbaccalaureate program designs), as well as the widely varying enrollment demographics that exist across the more than 1,000 universities and colleges that offer teacher certification programs. The goal of this research is largely descriptive: to understand better how teachers are being trained relative to current understandings of learning, teaching, and the development of expertise; how much variation currently exists in teacher education programs; and the factors that contribute to such variability. Of special concern are program structures, course content, and instructional practices that seriously conflict with the principles of How People Learn. The proposed research should also bring into focus features of teacher education programs that correspond to the principles of learning, and that enhance the capability of future teachers to incorporate the principles into their practice.

     9. Review professional development programs for alignment with the principles of learning and for relative effectiveness in changing teaching practice. The issue of teacher preparedness is rapidly becoming one of intense focus in policy arenas. Professional development programs are an important policy tool available to concerned lawmakers. But there are vastly different models of professional development, and relatively little is known about the amount and type that is required to significantly change teacher performance and student achievement. Existing research efforts along these lines need to be extended and built on.

     The committee recommends that alternative models of professional development be reviewed for their alignment with the principles of learning. Features that promote or conflict with the principles should be highlighted. The research should also examine the effects of alternative types, and amounts, of professional development training on teacher performance and student achievement. As envisioned, the research would:

     a. Define a small set of common models of professional development. These should include individual workshops, more lengthy in-service programs, and university courses. They should include training that is tied to a specific curriculum, as well as training in teaching techniques.

     b. Review the features of those programs that do and do not support learning, including the opportunities they provide for exploring teachers' preconceptions, for assessing what teachers are learning as they go along, and for teachers to provide feedback and receive ongoing support as they attempt to use what they have learned in the classroom environment.

     c. Define measures of teacher knowledge and performance that would be expected to change as a result of the learning opportunity.

     d. Define measures of student achievement that would be expected to change as a result of the change in teaching.

     e. Estimate the effect of quantity and type of training on teacher performance and student achievement.

     The envisioned research would require a major data collection effort. Success is likely to require that researchers work closely with school districts over a multiyear period. In states or school districts that are about to undergo an expansion in professional development spending, conditions may be particularly ripe for such a partnership.

     The results of this research should be written up separately for the three communities who are likely to find them useful: (a) for those who provide professional development programs, the results should provide feedback that allows for improvement in program design; (b) for administrators and policy makers, the results should provide guidance in evaluating professional development programs; and (c) for researchers, the results should be reported in detail sufficient to support further meta-analytic research.

     10. Explore the efficacy of various types of professional development activities for school administrators. School administrators at the individual school and school district level are responsible for facilitating teacher learning and evaluating teacher performance. If they are to support teachers' efforts to incorporate the principles of learning into classroom practice, they will need professional development opportunities that provide an understanding of the principles and their enactment in a classroom environment.

     The committee recommends that research be conducted to identify the amount and type of professional development needed to create in administrators an ability to differentiate between teaching practices that do, and do not, incorporate what is known about how people learn. This research should go beyond an effort to identify whether a particular professional development opportunity effectively changes administrators' evaluations of teacher performance. It should vary the amount of such training and the model through which training is provided (intensive workshops, monthly seminars conducted over the course of a year, etc.). Measures of administrators' interpretations of teaching should be taken prior to training, at the point of program completion, and again a year after completion in order to ascertain the sustainability of change over time and the effect of prior beliefs on post-training performance.

     11. Conduct research on the preconceptions of teachers regarding the process of learning. Adults, as well as children, have preconceptions that contribute to the ways in which they make sense of ideas and evidence and the decisions they make in undertaking tasks. For teachers to think about and conduct their teaching differently, they need to learn, and the principles of learning should guide that effort. The committee therefore recommends that:

     a. Research be conducted that explores the prior conceptions and beliefs of teachers and those learning to become teachers, identifying the common pedagogical models that current and prospective teachers use;

     b. Learning opportunities be developed that challenge misconceptions about how people learn and support the development of a new model that is based on learning research;

     c. Evaluations should be conducted of the effectiveness of those learning opportunities in changing understanding and conceptions of practice.

     The outcome of this research would include both a description of common preconceptions about learning, as well as tested techniques for working with those preconceptions that could be incorporated into the curricula of schools of education and professional development programs.

     12. Conduct discipline-specific research on the level and type of education required for teaching that discipline in elementary, middle, and high school. How People Learn makes clear that to teach effectively in any discipline, the teacher must link the information being taught to the key organizing principles of the discipline. To achieve this, the teacher must be provided with the discipline-specific training that allows for deep understanding of those principles. This type of teaching is not now a consistent feature of teacher training programs.

     The committee recommends that discipline-specific research be conducted on the amount and type of training in content knowledge that teachers need for various levels of schooling (elementary, middle, high) in order to teach for understanding. The challenge in providing such training is to equip the future teacher with both content knowledge and an understanding of the thinking of children in the subject area at different developmental stages. Each is a critical component for effective teaching in a subject area. In light of this dual requirement, is content knowledge best obtained in disciplinary courses that also service majors in the discipline, or in courses in schools of education, or in jointly sponsored courses that emphasize effective teaching of the content of the discipline? When content and teaching methods are taught separately, are teachers able to bridge the two? When they are done together, is adequate attention given to the disciplinary content?

     The committee recommends further that the discipline-specific research teams evaluate existing tools for assessing teachers' content knowledge and knowledge of discipline-specific developmental trajectories and make recommendations regarding their adequacy.

     13. Develop model pedagogical laboratories. In many fields in which scientific principles must be put to work, laboratory experiences provide the opportunity to experiment with applications of general and specific principles. The expense of the laboratories is justified by the qualitatively different experience made possible when the boundaries of an idea can be tested or worked with in a laboratory or field-based setting.

     To prepare students in schools of education to put to work the scientific principles of how people learn, laboratory experience could provide the opportunity to test the principles, become familiar with their boundaries, and learn how to make them operational. The committee therefore recommends the development of pilot pedagogical laboratories.

     The teachers who worked with the committee emphasized that a first classroom experience can so overwhelm a teacher that what was learned in a preparatory program can quickly be cast aside. Norms of operating in a school can quickly be adopted as survival techniques, however divergent those norms and the principles of learning might be. Laboratory experience could provide the opportunities for practice, as well as for observation and diagnosis of events that are likely to arise in the classroom, that could ease the transition into the classroom and allow for greater transfer of school-based learning to the practice of teaching.

     The laboratories, as envisioned, would have multiple purposes, the most important of which would be to provide teaching practice. The laboratories would need to develop ongoing relationships with a body of students to be taught (e.g., partnerships with local schools or Saturday classes). How this relationship would be established and maintained should be given careful attention in the design proposal for such a laboratory. Expert teachers who staff the laboratory would provide feedback and diagnosis of the teacher's lessons. The process could be aided by the use of a videotaped record of the instruction. The analysis could be further augmented by viewing tapes of other teachers who have attempted similar lessons. The teacher in training would work to improve the lesson through an iterative process of feedback and revision.

     The laboratory setting would be ideal for helping teachers to develop the ability to conduct formative assessment techniques. A theme that has coursed through this report is that teachers must be able to draw out and work with students' preconceptions and assess their progress toward understanding. The laboratory could provide opportunities to develop those techniques under guided instruction.

     The lab, as envisioned, would not provide a teaching internship or serve the function of a professional development school. Rather, it would provide an opportunity for beginning teachers to experiment with the principles of learning that are relevant to teaching practice. The goal is not to decontextualize teaching, but to create an environment in which the immediate demands of the classroom do not prevent reflection on, or exploration of, the process of learning. Exercises could be developed for laboratory use that involve cognitive science findings of relevance to teaching, including findings on memory, the organization of information, the use of metacognitive strategies, and retrieval of knowledge when transfer is prompted and when it is not. In addition to creating a deeper appreciation of the science of learning, these opportunities would invite teachers to think of themselves as scientists, to observe and reflect on learning as a scientist would. To the extent that those skills transfer to the classroom, the goal of continuous learning and reflection on practice will be well served.

     The laboratories would also serve as a locus of information for teachers in training, for practicing teachers in the community, and for researchers in the learning sciences. "Protocol materials," or materials for diagnosis and interpretation, could be housed here. These might include model lessons or units (project area 4) that could be incorporated into the teaching of diagnostic and interpretive competencies. They might also include protocols of student creativity in scientific thinking, insight, reasoning like a novice versus an expert in a task, failure to transfer, negative transfer, distributed cognition, using parental stores of knowledge in a class, concrete and operational thinking, and inferring causation. These protocols, then, provide vivid cases and examples that instantiate concepts relevant to teaching and learning. Videotaped lessons of teaching in other countries produced by the Third International Mathematics and Science Study project might also be made available. Faculty-directed course projects could develop evaluations of curricula in terms of the principles of learning and submit them to the interactive communications site described above (project area 7) for broad use.

     Technology centers could be housed in the laboratory as well. Computer programs to support classroom learning and technology-based curricula could be made available for exploration in this setting. Opportunities to connect with relevant communities of teachers and researchers via the Internet could also be explored. Students graduating from these programs will then carry to the schools in which they teach an ability to be connected to outside communities with relevant knowledge that is not now a feature in many school districts.

     Well-equipped laboratories would be an asset in professional development activities as well as in pre-service training. As such, the laboratories could be used on a year-round basis.

     14. Develop tools for in-service education that communicate the principles of learning in How People Learn. For the principles of learning to be incorporated into classroom practice, practicing teachers are a key audience. They are also a very busy audience. The challenge of developing ways to effectively communicate to those teachers is a central one. The committee recommends research and development that distills the messages of How People Learn for teachers and develops examples that are relevant to the classroom context. These messages should be communicated in a variety of formats, including text, audiotapes, videotapes, CD-ROMS, and Internet-based resources.

     Researchers should design and study the effectiveness of the different media in communicating key ideas, as well as the satisfaction of teachers with the various media and the change in practice that ensues. This research should focus on the format of the material as well. For example, case-like stories could be compared with more didactic methods often used in texts and lectures.

     How People Learn, we have argued, suggests far-reaching reform of education. It has direct implications for what is taught in the classroom, how it is taught, the relationship between students and teachers, the content and role of assessments, and the preparation of those who undertake the daunting task of classroom teaching. Yet How People Learn is not a blueprint for redesigning schools.

     Policy makers involved in the workshop were interested in the critical components of change that the report implies, as well as their associated costs. Given the task that is before them, this focus can be easily understood. But just as a doctor who recommends a healthy diet, stress reduction, exercise, adequate rest, and a personal support system cannot say which is most critical to health, researchers cannot identify the most critical change in the education system. The parts of the system cannot be isolated; the interactions among them have powerful influences on outcome.

     And just as the exercise requirement has no single attached cost--it can be met by a run through the park or an indoor tennis game at a posh racket club--teaching for understanding has no obvious price tag attached. Eliciting and working with student ideas and preconceptions will be easier in a small class than in a larger one, just as exercise in a sports club will be easier in inclement weather. But with a diverse clientele, a doctor will do best to focus on the principle of raising the heart rate for a sustained period of time rather than dictate the method for achieving the goal. Similarly, we focus on the principles of teaching for understanding with the recognition that, in the diverse landscape of schooling, the manifestations of those principles will vary. This does not diminish what is known with certainty: teaching for understanding is a clear goal with several well-defined components (discussed in Chapter 2).

     Our focus here is on policies that have a direct impact on attainment of those goals. Many of the research efforts already recommended will help inform policy; research on the efficacy of professional development programs, for example, will be of use to policy makers who set requirements for receiving funds for that purpose. At the urging of both policy makers and educators who participated in the workshop, we propose further research to review standards and assessments at the state level, and to examine teacher certification requirements at both the state and national level.

     At the district level, reform can be notoriously difficult to implement or extend. In order to identify the policies that appear to facilitate or impede the adoption and expansion of new teaching practices, we propose case study research on schools and school districts that have successfully implemented reform. Although we don't envision a blueprint, there may be organizational features, operational policies, or incentive structures in these schools that create an environment conducive to change.

     The recommended research is described in five project areas.

     15. Review state education standards and the assessment tools used to measure compliance through the lens of How People Learn. Forty-nine states now have a set of education standards that apply to their schools, and most have or are developing assessment tools to hold school districts accountable for implementation. Standards vary considerably in the amount of control they exercise over what is taught, in the content they impose, and (implicitly or explicitly) in the model of learning that they imply. The committee recommends that a sample of state standards be reviewed through the lens of How People Learn for the following purposes:

     a. To identify features of standards that support and violate the principles of learning in How People Learn;

     b. To evaluate the alignment of desirable features in a state's standards with the assessment tools used for measuring compliance;

     c. To evaluate the features of compliance assessments that support and conflict with the principles of learning;

     d. To identify incentives and penalties that support the goal of effective education and those that appear to undermine that goal.

     16. Conduct research on measures of student achievement that reflect the principles in How People Learn and that can be used by states for accountability purposes. Tests of student achievement that can be widely and uniformly administered across schools are the key mechanism by which policy makers hold schools accountable. How People Learn has clear implications for the measurement of student achievement. It suggests, for example, that recall of factual information is inadequate as a measure of deep understanding or as an indicator of the ability to transfer learning to new situations or problems.

     Conventional psychological and educational testing is an outgrowth of theories of ability and intelligence that were current at the beginning of the century. Psychometrics has become increasingly sophisticated in its measurements; yet it does not attempt to look inside the "black box" of the mind. Now that the newer sciences of cognition and development have transformed our understanding of learning and the development of expertise, measurement theory and practice need fundamental rethinking. There is much in the traditional methods that is valuable, including a focus on objectivity and reliability of measurement. There is a problem, however, with what is being measured.

     As a first step in the process of rethinking educational testing, the committee recommends that assessment tools be designed and tested with the goal of measuring deep understanding as well as the acquisition of factual knowledge. This is both a modest beginning and a challenging task. To be useful for policy purposes, these assessments should be in a form that can be administered widely and scored objectively and that meets reasonable standards of validity and reliability. These requirements can be at odds with the measurement of deep understanding, at least in the current state of the art. But it is important to begin finding solutions that, for example, minimize the trade-off between assessing for understanding and scoring objectively. A variety of experiments is needed, both with new forms of standardized tests (including computer-based instruments that permit "virtual" experiments), and with alternative assessments (such as portfolios) that have become more popular in recent years.

     The committee further recommends research on assessment tools of different types to determine:

     a. Whether alternative assessments yield significantly different measures of student achievement or highly correlated results.

     b. How alternative assessment measures might be combined to offer a balanced view of achievement.

     17. Review teacher certification and recertification requirements. Currently, 42 of the 50 states assess teachers as part of the certification and licensure process. But states vary enormously in the criteria used and the amount and type of assessment they require. The federal government also has provided support for an assessment process for advanced certification that is developed and administered by the National Board for Professional Teaching Standards. The committee recommends that research be conducted to review the requirements for teacher certification in a sample of states (selected for their diversity). Specific focus should be given to the types of assessments currently in use across the continuum of teacher development, from initial licensure to advanced status. This would include standardized tests, performance-based assessments under development (Interstate New Teacher Assessment and Support Consortium), and the National Board assessments. Efforts should be made to determine:

     a. The features of certification that are aligned with the principles of How People Learn and those that are in conflict.

     b. To the extent that data are available, the relationship between certification and increases in student learning.

     This project should also recommend, when appropriate, strategies to reform certification processes so that they provide better signals of a teacher's preparedness for the task of teaching for understanding.

     18. Conduct case study research of successful "scaling-up" of new curricula. School districts set a variety of policies that influence the environment in which teachers operate. Even when a new curriculum is pilot-tested with positive results, it can be very difficult to extend that curriculum into other schools in the district, sometimes even to other classrooms in the same school. The committee recommends case study research of successful scaling-up efforts to determine which district-level and school-level policies facilitated reform. The case studies should include information on features that teachers often identify as obstacles to reform:

     a. How much scheduled time do teachers have in their work day that is not in the classroom and that can be used for reflection, study, or discussion with other teachers?

     b. How much training was offered to teachers who adopted the new curriculum? Is there ongoing support for the teacher who has questions during implementation? Is there evaluation of the teacher's success at implementation?

     c. Is there a community within the school, or extending beyond the school, that provides support, feedback, and an opportunity for discussion among teachers? Existing research suggests that the development of a professional community as part of the school culture is one of the most important determinants of successful school restructuring to implement a more demanding curriculum (Elmore, 1995; Elmore and Burney, 1996). These studies should focus on the features that hold that community together. Are there key players? Are there structured or informal opportunities for the exchange of ideas? What can be learned from these successes about the opportunities for enhancing teacher access to communities of learning using Internet tools?

     d. Did the school attempt to involve parents and other community stakeholders in the change?

     Some case study research of this type has already been done or is now under way. The effort to extend the knowledge base in this area should be coupled with an effort to synthesize the research results, making them easily accessible to school communities interested in reform.

     19. Conduct research on the effective communication of research results to policy makers. Policy makers do not routinely look to research as a source of information and ideas. But there are windows of opportunity for research in policy making. Researchers who study this issue suggest that the windows are more likely to open during crises, when issues are new and policy makers have not yet taken a position, or when issues have been fought to a stalemate. When those opportunities arise, information must be communicated to policy makers in a manner that optimizes the chance that they will learn from research findings.

     The committee recommends that research be conducted to:

     a. Assess preconceptions of education policy makers regarding the goals of K-12 education and the strategies for achieving those goals. Are they consistent with the principles of learning in How People Learn?

     b. Identify examples that engage the preconceptions of policy makers (if those preconceptions diverge from research findings on how people learn) and test their effectiveness at changing the initial understanding.

     c. Identify methods of communication that are most likely to reach, and teach, policy makers.

     d. Compare the effectiveness of alternative approaches, including concisely written materials, personal contact, and briefings or seminars.

     The product of this research should be both a report of the findings regarding how policy makers learn most effectively and concisely written material that can be used for communicating effectively to policy makers.

     Information communicated to the public through the media can influence practice in two ways. First, to the extent that the public is aware of the implications of learning research for classroom practice, teachers, administrators, and policy makers will receive more support for the types of changes that How People Learn suggests. Second, many teachers, administrators, and policy makers themselves are influenced by ideas that reach them through popular media. As we heard from participants in the workshops, How People Learn is not a document that is likely to be widely read by educators and policy makers. Information presented in a more popular format will have far better prospects of reaching this audience.

     20. Write a popular version of How People Learn for parents and the public. Everyone has preconceptions regarding the process of learning and effective methods of education. Those theories are put to work on a daily basis when we model behaviors for children, provide instructions to coworkers, or explain a problem to a friend. These models are likely to be influenced by personal experience.

     The translations of these experience-based models to the evaluation of classroom teaching can lead to expectations that conflict with the principles of learning drawn from research. A parent who is accustomed to teaching a child through direct instruction, for example, may be baffled by mathematics homework that requires the child to find a method of adding five two-digit numbers, rather than instructing the child to line those numbers in columns and add the columns in turn. The importance of grappling with the problem and searching for a solution method, and the appreciation that such grappling brings to the conventional method of solution, can be lost on the parent.

     How People Learn develops many concepts and ideas that could inform parents about models of learning that are research based, thus influencing the criteria that parents use to judge classroom practice. But those ideas are embedded in a report that is not designed specifically to communicate to parents. The committee recommends the writing of a popular version of How People Learn. The popular presentation should address common preconceptions held by the public regarding learning. It should couch research findings in multiple examples that are relevant to parents' observations of children at a variety of ages. And it should help parents who are interested in understanding or evaluating a school formulate questions and make observations.

     Some particularly effective examples and their implications for teaching should be highlighted in a manner that makes them easy to extract from the text. The children's book, Fish is Fish by Leo Lionni (1970) mentioned in How People Learn, can serve as an effective example. In the story, a frog adventures onto the land and comes back to describe what it saw. The fish who listen to the frog imagine each description to be an adaptation of a fish: humans are imagined to have fish bodies but walk upright, etc. The visual image powerfully describes the problem of presenting new information without regard to the learner's existing conceptions. Examples such as these would allow the popular media to communicate key ideas to the broader public who might not read the report.

     The popular version of the report should itself be a subject of study. The committee proposes that a second stage of this project should involve research to assess whether the popular report effectively communicates its messages to a sample of parents.

     The research and development agenda proposed thus far focuses on the question that the Office of Educational Research and Improvement posed to the committee: How can the insights from How People Learn be incorporated into educational practice? How People Learn reviews a burgeoning literature that, taken collectively, provides the foundation for a science of learning. But more work needs to be done to extend that foundation. The committee proposes three research projects that hold promise for advancing the understanding of learning in ways that will be useful for educational practice.

     21. Investigate successful and creative educational practice. There are well-known cases of exceptional teaching by educators who, often without the help of educational researchers, have created innovative and successful classrooms, programs, curricula, and teaching techniques. The committee recommends that case study research be conducted to investigate the principles of learning that underlie successful educational experiments. The conceptual framework provided by How People Learn can be employed as a lens through which that practice can be viewed, and such case studies could challenge and inform the science of learning.

     The research would have several potential benefits: it would ground in sound theory innovations that often exist in isolation, that often cannot be evaluated well by traditional methods, and that cannot be explained well to others. This research could contribute an understanding of why the innovations work, perhaps leading to improvements in them. Moreover, it may stimulate researchers to pursue new theoretical questions regarding cognition. In innovative classrooms, students may engage in forms and levels of learning that are not anticipated by current cognitive theory. From studying such classrooms and the learning that takes place in them, researchers may modify their conceptions about learning.

     22. Investigate the potential benefits of collaborative learning in the classroom and the design challenges that it imposes. Outside the classroom, much learning and problem solving takes place as individuals engage with each other, inquire of those with skills and expertise, and use resources and tools that are available in the surrounding environment. The benefits of this "distributed cognition" are tapped inside the classroom when students work collaboratively on problems or projects, learning from each others' insights, and clarifying their own thinking through articulation and argument (Vye et al., 1998a). Some research indicates that group problem solving is superior to individual problem solving (e.g., Evans, 1989; Newstead and Evans, 1995), and that developmental changes in cognition can be generated from peer argumentation (Goldman, 1994; Habermas, 1990; Kuhn, 1991; Moshman, 1995a, 1995b; Salmon and Zeitz, 1995; Youniss and Damon, 1992), and peer interaction (Dimant and Bearison, 1991, Kobayashi, 1994). For these reasons, the community-centered classroom described in Chapter 2, in which students learn from each other, can have substantial benefits.

     But working in groups can have drawbacks for learning as well, particularly in the early grades. Societal stereotypes or classroom reputations can determine who takes the lead, and whose ideas are respected or dismissed. Differences in temperament can produce consistent leaders and followers. Group products can advance each members' understanding of a problem, or they can mask a lack of understanding by some.

     The committee recommends that research be conducted by teams of cognitive scientists, developmental psychologists, curriculum developers, and teachers to investigate the potential benefits of collaborative learning in the classroom and the problems that must be addressed to make it beneficial for all students. The research should explore and field-test alternative design strategies. The results should be presented both as scholarly research, and as a discussion addressed to teachers who are interested in collaborative learning in the classroom.

     23. Investigate the interaction between cognitive competence and motivational factors. Much of the research on learning has been conducted outside the classroom. Inside the classroom, issues of cognitive competence are intertwined with issues of motivation to perform. The challenges of learning for today's world require disciplined study and problem solving from the earliest grades. To meet the challenges, learners must be motivated to pay attention, complete assignments, and engage in thinking.

     Although cognitive psychologists have long posited a relationship between learning and motivation, they have paid little attention to the latter, despite its vital interest to teachers. Research has been done on motivation, but there is no commonly accepted unifying theory, nor a systematic application of what is known to educational practice (National Research Council, 1999b).

     The committee recommends that research be conducted to elucidate how student interests, identities, self-knowledge, self-regulation, and emotion interact with cognitive competence. This research should combine the efforts of social and developmental psychologists with those of cognitive psychologists. A variety of approaches should be considered, including case studies of small numbers of individual children and the study of the classroom practice of teachers with reputations for promoting achievement among average students, as well as those at high risk for failure.

     24. Investigate the relationship between the organization and representation of knowledge and the purpose of learning that knowledge. Research in cognitive science suggests that knowledge is organized differently depending on the uses that need to be made of it. In other words, the structure of knowledge and memory and the conditions under which it is retrieved for application evolves to fit the uses to which it is put. Similarly, what counts as understanding will also be defined in terms of means, rather than as an end in itself. Just as there is no perfect map, but only maps that are useful for particular kinds of tasks and answering particular kinds of questions, there is no perfect state of understanding, but only knowledge organizations that are more or less useful for particular kinds of tasks and questions.

     For example, relatively superficial knowledge of the concept of gold may be sufficient to differentiate a gold-colored watch from a silver-colored watch. But it would not be sufficient to differentiate a genuine gold watch from one made of other gold-colored metals or alloys, or fool's gold from the real thing.

     This empirical insight has profound implications for the organization of education, teacher education, and curriculum development. The committee recommends research to deepen understanding of the kinds of knowledge organizations that will best support particular kinds of activities. For example, the kinds of biology needed to know how to take care of plants (e.g., knowing when, where, and how to plant them in different climates and soil conditions) differs from the knowledge necessary to genetically engineer them.

     These kinds of issues become particularly important when considering the nature of the content knowledge that teachers need in order to teach various disciplines. For example, the most useful knowledge for a middle school mathematics teacher may not come from taking a higher-level course in a traditional mathematics sequence, particularly if that course was designed for the uses of that knowledge by mathematics and engineering students in problems suited to the work activities of those disciplines. Instead, it may come from a course that integrates mathematics with particular kinds of inquiry involving design and other tasks.

     These considerations are also important for curriculum. Research investigations could yield better understanding for guiding curriculum design so that the knowledge that learners develop from their experiences in courses will be better retrieved in anticipated contexts of use for that knowledge. For example, too little is known about the kinds of activities in which an educated person--but not a future scientist--will be expected to use the scientific knowledge that they may acquire in science courses. Research on these considerations is important to pursue.

     When one considers the complexity of the ways in which research influences practice (as depicted in Figure 1.1), the heterogeneous audiences for research and their very different needs become apparent. As we have noted throughout this report, the ways in which the principles of learning depicted in How People Learn will be incorporated into practice raise unique problems for pre-service and in-service education, for educational materials, for policy, and for the public (including the media). The pathways by which research knowledge travels, and the transformations it must undertake for each of these audiences, raise striking challenges for communications design. To be effective, such communications cannot serve merely as disseminations of research knowledge. Translating and elaborating that knowledge for each audience has been a theme throughout the agenda. In this final section, we propose an effort to make these translations widely accessible.

     25. Design and evaluate ways to easily access the cumulative knowledge base. Adaptive communications about the science of learning are very much needed that can evolve to fit the distinctive needs of the various education audiences for knowledge derived from research. For such conversations to occur between the research communities and these diverse constituencies, experimentation with Internet-based communications forums is needed.

     The Internet is becoming a social place for the formation and ongoing activities of distributed communities, not only a digital library for browsing and downloading information. Current electronic communities with tens of thousands of members share information and convene around a broad range of topics. High-quality resources on the science of learning will be needed to spur on-line discussions among the communities they are designed to serve, and to invite suggestions about how communications concerning the science of learning can better fit the needs of those who will use their results (Pea, 1999). Today one may find a great range of web sites that are devoted to education. But far fewer are devoted to research advances, much less their alignment with educational materials, practices, or policies that are depicted in the web sites.

     The committee recommends the development and continuous improvement of a national communications forum for research knowledge on learning and teaching. This new media communications forum would be accessible through the Internet and would provide illustrative cases and usable information about both the research depicted in How People Learn and new findings that will continue to emerge in ongoing research. It would provide opportunities for different contributors who are stakeholders in education to post messages and rate the usefulness of documents and materials. Experimentation is needed in establishing "virtual places" online where diverse groups could convene to reflect on how these research advances could be incorporated to improve the practices of education and learning. Such a "learning improvement portal" would provide a vital national resource, guiding research-informed improvements of education.

     The research efforts that we have proposed make a serious effort to combine the strengths of the research community with the insights gained from the wisdom and challenges of classroom practice. Our suggestions for research do not assume that basic research should first be conducted in isolation and then handed down to practitioners. Instead, we propose that researchers and practitioners work together to identify important problems of inquiry and define the kinds of research and communication strategies that would be most helpful to both groups.

     Because of our emphasis on bridging research and practice, many of the efforts that we have proposed here are nontraditional. They combine research and development, rather than undertaking the two separately. It is the committee's view that such combined efforts are most likely to focus the attention of researchers on problems that are central to education, and they are more likely to ensure rigor and consistency with the principles of learning in the programs and products that are developed.

     Moreover, many of the efforts combine research and communication. Often, the two are considered separate domains. But the goal of communication is learning, and How People Learn provides guidance for effective communication. For each audience, preconceived understandings must be identified and addressed in the effort to communicate. And examples that situate ideas in experiences relevant for that audience are crucial.

     Combining expertise for the proposed projects will be challenging. There are still relatively few arenas in which researchers work as partners with teachers, administrators, and communications developers (who might film model lessons, develop web sites, produce brochures, etc.). But to be effective, systematic efforts to reform education will require that more of these partnerships be forged. Research and development grants that reward existing partnerships and encourage new ones to be formed could provide a much-needed impetus.

     And finally, the agenda proposed is expansive. Many of the recommended projects are time-intensive, multiyear efforts. The nation's decentralized education system is vast. To don the lens of How People Learn to evaluate the various facets of that system is in itself a daunting task. We propose further the development and testing of new classroom teaching tools, techniques of teacher and administrator training, further research on human learning, and applications of technology that could provide dynamic mechanisms for bringing advances in how people learn and how people teach into continual cycles of coordination and improvement. From the committee's perspective, the integration of these efforts holds the potential to bring research and practice together in the interest of improved education.