Conclusions and Recommendations
The emerging understanding of how children learn science in kindergarten through eighth grade paints a very different picture of a science learner than existed 20 or 30 years ago. As this report has documented, children come to school with rich knowledge of the natural world and an ability to engage in complex reasoning that provides a solid foundation for learning science. At the same time, many key ideas and ways of thinking in science are difficult if not impossible to achieve without instructional support. Successful strategies for science learning engage students in scientific tasks that explore ideas and problems that are meaningful to them with carefully structured support from teachers. Too often, however, the instructional and curricular approaches currently used in classrooms do not reflect this emerging understanding of children as competent learners who can engage in scientific tasks throughout their schooling. Instead, current approaches are often based on now outdated knowledge about cognitive development and misunderstanding of its implications concerning how to design instruction for young and novice learners.
In this chapter, the committee summarizes the major conclusions of the report. We then follow with a discussion of the key recommendations for policy and practice that flow from these conclusions. Finally, we outline a research agenda that if pursued would fill critical gaps in the knowledge base and recommends a multidisciplinary approach to the issues that have emerged in the report.
MAJOR FINDINGS AND CONCLUSIONS
We begin with a discussion of current understanding of science learners and science learning, highlighting the ideas that differ from some popularly held conceptions. These conclusions are based on evidence discussed mainly in Part II. Working from this picture of learning, we discuss what is known about effective curriculum and instruction. These conclusions flow from the new vision of learning described in Part II as well as from studies reviewed in Part III that have explicitly explored the design of effective science instruction. Finally, we move from the classroom into the larger context of the school and district to consider key factors the committee thinks influence whether and how classroom practice is informed by knowledge about how children learn science. Evidence for these conclusions is discussed mainly in Chapter 10.
Learning and Learners
The committee developed a framework for proficiency that identifies four fundamental strands of learning. In our view students who understand science:
Know, use, and interpret scientific explanations of the natural world.
Generate and evaluate scientific evidence and explanations.
Understand the nature and development of scientific knowledge.
Participate productively in scientific practices and discourse.
These strands of scientific proficiency represent learning goals for students as well as provide a broad framework for curriculum design. They address the knowledge and reasoning skills that students must eventually acquire to be considered fully proficient in science. They also incorporate the scientific practices that students need to participate in and become fluent with in order to demonstrate their proficiency. Students can understand science with increasing sophistication starting in the earliest grades. The committee thinks the development of proficiency is best supported when classrooms provide learning opportunities that interweave all four strands together in instruction.
Evidence to date indicates that in the process of achieving proficiency in science, the four strands are intertwined so that advances in one strand support and advance another. For example, learning how to design controlled experiments enables students to discover and verify knowledge about causal factors in the natural world. Conversely, an inadequate understanding of how scientific knowledge is constructed can constrain students’ scientific reasoning and limit the kinds of inferences they may be drawn from evidence.
This four-strand framework represents an important departure from the dichotomy of content and process that informs much of current practice in science education. That is, teaching content alone is not likely to lead to proficiency in science, nor is engaging in inquiry experiences devoid of meaningful science content. In current practice, content and an oversimplified view of scientific processes are often the primary or even sole foci of instruction; however, the evidence indicates that this leads to a very impoverished understanding of science and masks the complex process involved in developing scientific evidence and explanations. In addition, without an understanding of how scientific evidence is obtained, evaluated, and accumulated—more sophisticated than the seven-step version of the scientific method that is often taught in U.S. schools—students are unlikely to become science-literate citizens who can critically evaluate scientific information in order to participate in public debates or make informed decisions. Students are most likely to be successful in science when the four strands are brought together in instruction.
Conclusion 1: The norms of scientific argument, explanation, and the evaluation of evidence differ from those in everyday life. Students need support to learn appropriate norms and language for productive participation in the discourses of science.
Many advances in children’s knowledge and understanding of science are made through a social process of constructing arguments and evaluating evidence. However, the rules for engaging in arguments and evaluating evidence that students learn in their everyday lives are sometimes dissimilar and even contradictory to those employed in science. For example, children may settle arguments outside the science classroom on the basis of appeals to authority, personal experience, social status, or physical size. In contrast, children engaged in a productive scientific argument base their claims on empirical evidence and engage in arguments in order to refine their thinking and clarify their collective understanding of phenomena. Students often need support or explicit guidance to learn scientific norms for interacting with peers as they argue about evidence and clarify their own emerging understanding of science and scientific ideas.
Conclusion 2: Children entering school already have substantial knowledge of the natural world, much of it implicit. In contrast to the commonly held and outmoded view that young children are concrete and simplistic thinkers, the research evidence now shows that their thinking is surprisingly sophisticated. They can use a wide range of reasoning processes that form the underpinnings of scientific thinking, even though their experience is variable and they have much more to learn.
Recent research indicates that children of all ages can and do engage in complex reasoning about the world. This echoes a major conclusion of How People Learn: Brain, Mind, Experience and School, Expanded Edition (National Research Council, 2000, National Academy Press, Washington, DC) that “children lack knowledge and experience, but not reasoning ability,” that is, “they are ignorant, but not stupid” (p. 112). Children reason quite well with the knowledge they do understand. Furthermore, it is now known that even preschool and kindergarten-age children have a more sophisticated knowledge of the natural world than was once assumed.
Much of the current science education curriculum is based on dated assumptions about the nature of cognitive development and learning, assumptions that lead to the suboptimal teaching of science. It has been common to view younger children as deficient in some manner, resulting in a focus on what they cannot do rather than what they can do. Cognitive development has often been characterized as a series of artificial dichotomies, in which children do or do not have a particular capacity and the transition from not having the capacity to having the capacity is understood as going through a stage. These deficit assumptions ignore the tremendous subtlety, variability, and context dependence in children’s thinking and reasoning, and the important domain-specific knowledge they bring to school, especially knowledge of the natural world. In many cases, these assumptions about children’s abilities lead to curriculum materials that are designed to fit with a stage-like conception of a child’s abilities and inabilities, rather than to take advantage of and build on existing knowledge and reasoning skills.
Conclusion 3: What children are capable of at a particular age is the result of a complex interplay among maturation, experience, and instruction. Thus, what is developmentally appropriate is not a simple function of age or grade. What children can do is in large part contingent on their prior opportunities to learn and not on some fixed sequence of developmental stages.
Contrary to conceptions of development held 30 or 40 years ago, current research does not show a broad age trend in children’s thought suggesting that young children are only able to think concretely, with abstract thought emerging only in later childhood. Instead, there is variation across children at a given age and even variation within an individual child. That is, a single child’s thinking does not develop in a unitary way across all domains; at a given point in time, a child may be more sophisticated in one area and less sophisticated in another.
In addition, current research contradicts the assumption that development is a kind of inevitable unfolding and that one must simply wait until a child is cognitively “ready” for more abstract or theory-based forms of con-
tent. Instead, children need assistance to learn; building on their early capacities requires catalysts and mediation (National Research Council, 2000). Adults play a central role in “promoting children’s curiosity and persistence by directing their attention, structuring their experiences, supporting their learning attempts, and regulating the complexity and difficulty of levels of information for them” (National Research Council, 2000, p. 235). In the case of the science classroom, both teachers and peers can and must fill these critical roles.
A major problem with assuming children’s capacity for sophisticated reasoning will unfold with minimal support is that what they are capable of doing without instruction may lag considerably behind what they are capable of doing with effective instruction. In fact, there is more information from research about starting points than about children’s potential for developing scientific proficiency under effective instructional conditions. There are very few examples of what students may be capable of by the end of eighth grade if they experience effective science instruction from the time they enter school.
Conclusion 4: Students’ knowledge and experience play a critical role in their science learning, influencing all four strands of science understanding. Children’s concepts can be both resources and barriers to emerging understanding. These concepts can be enriched and transformed by appropriate classroom experiences. Science learners require instructional support to engage in scientific practices and to interpret experience and experiments in terms of scientific ideas.
Children’s rich but naïve understandings of the natural world can be built on to develop their understandings of scientific concepts. Some areas of knowledge may provide more robust foundations to build on than others, because they appear very early and appear to have some universal characteristics across cultures throughout the world. Young children, even infants, track a wide range of relational and causal properties of the world around them. They tend to identify regularities in the world around them that can be linked to broad domains, such as physical mechanics and the living world. Various aspects of scientific thinking and investigation are also closely tied to students’ understanding of the natural phenomena being considered. For example, students’ beliefs about the natural world shape the hypotheses they choose to pursue and the investigations they design to test them.
Children’s understandings of the world sometimes contradict scientific explanations. These conceptions about the natural world can pose obstacles to learning science. However, their prior knowledge also offers leverage points that can be built on to develop their understanding of scientific concepts and their ability to engage in scientific investigations. Thus, children’s
prior knowledge must be taken into account in order to design instruction in strategic ways that capitalize on the leverage points and adequately address potential areas of misunderstanding.
Some aspects of modern scientific understanding are so counterintuitive and “unnatural” that a child is highly unlikely to arrive at that understanding without explicit instruction, for example, understanding atomic-molecular theory, plate tectonics, or genetics. There are also aspects of scientific thinking in which adults still demonstrate difficulty and require support to learn. These include differentiating theory and evidence, evaluating evidence that contradicts prior beliefs, and understanding how scientific knowledge is constructed.
To move to greater proficiency, students need help with learning how to engage in science and with connecting their own ideas to scientific explanations of the natural world. Thus, topics and specific investigations within those topics must be selected with great thoughtfulness and care, and the structure and sequence of those activities must be carefully planned. Young and novice students are likely to profit from study in areas in which their personal, prior experience with the natural world can be leveraged to connect with scientific ideas. They will also need teacher assistance to engage in and pursue fruitful scientific investigations.
Studies of instructional interventions carried out over weeks or months indicate that, with opportunities to practice or explicit instruction, even elementary and middle school children can master difficult concepts in science. However, to be successful, students need carefully structured experiences, scaffolded support from teachers, and opportunities for sustained engagement with the same set of ideas over extended periods of time (weeks, months, even years).
Current approaches to adjusting science instruction for young and novice learners may actually be counterproductive. For example, limiting them to learning about discrete science facts without opportunities for discussion, reflection, or direct investigation of the phenomena can lead to a very impoverished understanding of the ideas. Developing expertise in science means developing a rich interconnected set of concepts (a knowledge structure) that moves closer and closer to resembling the structure of knowledge in the science discipline. Memorizing lists of established scientific facts does not provide the kind of engagement with ideas that will produce rich, interconnected knowledge and reasoning.
Conclusion 5: Proficiency in science involves having knowledge of facts and concepts as well as how these ideas and concepts are related to each other. Thus, to become more expert in science, students need to learn key ideas and concepts, how they are related to each other, and their implications and applications within the discipline. This entails a process of
conceptual development that in some cases involves large-scale reorganization of knowledge and is not a simple accumulation of information. Such deep conceptual change is achieved more successfully when students receive instruction that integrates the four strands.
The difference between students who are less or more proficient in science is not only that the latter know more discrete facts. Instead, gains in proficiency often consist of changes in the organization of knowledge, not just the accretion of more pieces of knowledge. Learning an unfamiliar concept requires students to come to understand the concept’s appropriate implications and applications. Moreover, enough of the surrounding conceptual or theoretical framework must be in place and understood so that students’ interpretation of the new concept will be appropriately constrained. When students develop a coherent understanding of the organizing principles of science, they are more likely to be able to apply their knowledge appropriately and will learn new, related material more effectively. Knowledge of the salient factual details is necessary but not sufficient for developing an understanding of the discipline and its core ideas and principles.
Conceptual development can occur in many different ways, and some conceptual changes are more challenging than others. For example, when children develop commonsense frameworks that deviate substantially from those proposed by scientists, a considerable amount of conceptual work is required to achieve knowledge restructuring. Part of the difficulty of learning a new concept is letting go of a familiar but incorrect set of ideas. Major changes in conceptual frameworks are often difficult to grasp because they require learners to break out of their familiar frame and reorganize a body of knowledge, often in ways that draw on unfamiliar ideas. Making these changes is facilitated when students engage in metacognitively guided learning, when teachers use a variety of techniques (such as bridging analogies, thought experiments, and imagistic reasoning) to help students construct an understanding of new concepts, and when students have opportunities to strengthen their understanding of the new ideas through extended application and argumentation.
Learning science is often characterized as increasing what one knows about concepts, ideas, and issues associated with science-related topics. Science in school is often presented as a rather flat, nonhierarchical list of unrelated concepts. This approach is driven by an assumption that the simple accumulation of ideas or facts increases knowledge or understanding. This narrow construal of knowing science as simply knowing facts or understanding a specific causal mechanism can lead to underestimating the rich knowledge of the natural world children bring to school. It also may cause teachers to ignore the sophisticated cognitive capacities that children have available to them that can allow them to build new knowledge. Moreover, it
underestimates what it takes for students to be able to go beyond simply repeating the memorized facts to understanding their implications in contexts beyond those in which the ideas were originally encountered.
Conclusion 6: Race/ethnicity, language, culture, gender, and socioeconomic status are among the factors that influence the knowledge and experience children bring to the classroom. This diversity offers richness and opportunities in the classroom, and it also affects the kinds of support children need to learn science.
The challenge of helping all students achieve proficiency in science is daunting in the context of an increasingly diverse student population and persistent gaps in science achievement. Children’s experience varies with their cultural, linguistic, and economic background. For example, cultural differences in discourse patterns are well documented, and some students’ norms for discussion and social interaction may actually be at odds with the norms in scientific practice. Such differences mean that students arrive in the classroom with varying levels of experience with science and varying degrees of comfort with the norms of scientific practice.
It may be hard for teachers to recognize the strengths that diverse learners bring to the science classroom. For example, differences among students in norms for discourse, lack of familiarity with scientific terms, or limited proficiency in English may produce the impression that some students are unable to be successful in science. However, all students bring basic reasoning skills, personal knowledge of the natural world, and curiosity, which can be built on to achieve proficiency in science. Capitalizing on these resources requires teacher sensitivity to cultural and other background differences and willingness and skill to adjust instruction in light of these differences. Adjusting for variation in students’ background and experience does not mean dumbing down the science curriculum or instruction provided to certain groups of students, for example, by reducing science to sheer memorization of facts and terms.
Curriculum and Instruction
Conclusion 7: Many existing national, state, and local standards and assessments, as well as the typical curricula in use in the United States, contain too many disconnected topics given equal priority. Too little attention is given to how students’ understanding of a topic can be supported and enhanced from grade to grade. As a result, topics receive repeated, shallow coverage with little consistency, which provides a fragile foundation for further knowledge growth.
Comparisons of science standards and curricula in the United States with that of countries that perform well on international science tests reveal overly broad and superficial coverage of science topics in U.S. classrooms, with little attention to building links across concepts or developing a specific concept over successive grades. Furthermore, national standards documents in the United States have a circular pattern, in which almost all concepts are covered at every grade level. This is in contrast to high-achieving countries that follow a more “spiral” curriculum, in which more challenging concepts are introduced gradually over successive grades.
Science textbooks suffer from similar problems. They tend to cover many more topics than those used in high-achieving countries. Close analysis of middle school textbooks indicates that science topics are presented as a list of unrelated items with little or no regard to the relations among them. Textbooks and the accompanying classroom activities are not consistently framed around the central ideas in the disciplines. Such organization of both standards and curricula does not match what is known about how best to facilitate student learning.
Conclusion 8: Sustained exploration of a focused set of core ideas in a discipline is a promising direction for organizing science instruction and curricula across grades K-8. A research and development program is needed to identify and elaborate the progressions of learning and instruction that can support students’ understanding of these core ideas. The difficult issue is deciding what to emphasize and what to eliminate.
Findings from research about children’s learning and development can be used to map learning progressions in science. That is, one can describe the successively more sophisticated ways of thinking about a topic that can follow and build on one another as children learn about and investigate a topic over a broad span of time (e.g., 6 to 8 years). Steps in these progressions are constrained by children’s knowledge and skill with respect to each of the four strands. Reaching the hypothetical steps described in the progressions is also dependent on teachers’ knowledge and the effectiveness of their instructional practices.
The emerging research on learning progressions offers insight into how curricula, and accompanying systems for assessment, might be reorganized to better support science learning. For example, such sequences could be organized around a few central ideas in science that would be studied and developed in depth and at increasing levels of complexity across succeeding grade levels. They would be anchored at one end by what is known about the reasoning of students entering school and at the other end by expectations about what society wants middle school students to understand about
science. The best candidates for the scientific ideas on which to build learning progressions are those that are central to a discipline of science, are accessible to students in some form starting in kindergarten, and have potential for sustained exploration across grades K-8.
Conclusion 9: Students learn science by actively engaging in the practices of science. A classroom environment that provides opportunities for students to participate in scientific practices includes scientific tasks embedded in social interaction using the discourse of science and work with scientific representations and tools. Each of these aspects requires support for student learning of scientific practices.
The view of science as practice is emerging from research on the work of scientists as well as from research on student learning in the classroom. In this view, theory development and reasoning are components of a large ensemble of activities that includes conducting investigations; networks of participants and institutions; specialized ways of talking and writing; modeling using mechanical, mathematical, or computer-based models; and development of representations of phenomena. To develop proficiency in science, students must participate in the full range of practices.
Conclusion 10: Frequently, K-8 classroom investigations treat data collection and analysis as the end game of science. Instead, science is, and should be presented as, a process of building theories and models, checking them for internal consistency and coherence, and testing them empirically. Method should follow from theory, therefore presentation of scientific methodology should consider a broad range of methods, including acquiring and interpreting observational data and modeling (an important research method in astronomy, the earth sciences, and evolutionary biology), which should be reflected in the experiences provided in K-8 classrooms.
Current science education tends to overemphasize experiment relative to the wide array of forms of scientific investigation that are in use in the sciences. Although experimentation is one fundamental form of investigation in science, it is by no means the sole or definitive means. An overemphasis on recipes for data collection procedures—whether experimental, observational, or archival—may strengthen the misconceptions that some students hold about the so-called scientific method—the image that scientific discoveries emerge unproblematically if one just faithfully follows the steps outlined in the science text. The tendency for science in school to use highly structured investigations often aimed at verifying established scientific principles exacerbates this problem.
Forms of descriptive science that rely on planned and structured observation and modeling are important ways to conduct science that are also accessible to elementary and middle school students. Moreover, some sciences, such as evolutionary biology and geology, rely heavily on historical reconstruction, a form of scientific inquiry that is underrepresented in school science.
Conclusion 11: The artificial dichotomy that pits teacher-directed instruction against discovery learning is not productive. A range of instructional approaches is necessary as part of a full development of the four strands. All students need to experience these different approaches.
Instruction occurs in sequences of designed, strategic encounters between students and science. Any given unit of study may include episodes that are highly teacher-directed as well as structured student-led activities. Across time, quality instruction should promote a sense of science as a meaningful process of building and improving knowledge and understanding. Students may generate researchable questions, design methods of interrogating these, conduct data analysis, and debate interpretations of data. Individual lessons may also focus on discrete questions, concepts, facts, or methods of investigation.
Instruction can make science processes more explicit for learners. Instruction may illustrate for students how to engage in science processes, learn to do them more effectively, and develop better understanding of the content they are investigating. Instructional supports can be designed with conceptual models or dynamic simulations that make science concepts more transparent for learners, helping them bridge from their prior understandings to more sophisticated scientific understandings.
Sustained investigations serve as a way of sidestepping the common tendency to treat content and process as separable goals. In these investigations, students both develop knowledge and explanations of the natural world and generate and interpret evidence. These investigations must be carefully selected to link to important scientific ideas. If not, hands-on investigations can turn into mindless, fun activity with little connection to important ideas in science.
Designing these contexts requires careful attention to learning goals and instructional support for engaging in the practices of science. Forms of support that have been effective include highlighting the structure of scientific tasks, modeling and shaping scientific discourse, and encouraging students to articulate and reflect on both the process and products of investigation. Without support, students may have difficulty in finding meaning in their investigations, or they may fail to see why and how they are relevant to their other ongoing work in the science classroom.
Conclusion 12: Ongoing assessment is an integral part of instruction that can foster student learning when appropriately designed and used regularly. Assessments, whether formative or summative, need to be responsive to the full range of proficiencies that are implied by the strands. Assessment needs to be aligned with the research on students’ thinking as well as informed by the structure of the subject matter.
Planning, evaluating, and improving the quality of science instruction is contingent on accurately assessing students’ knowledge and skills and how these develop over time. Individual teachers can conduct assessment to gauge student learning through the activities they use regularly in the course of instruction (e.g., questioning strategies, discussion, analysis of student work). Schools and school systems can administer periodic benchmarking assessments to track student learning over time and provide teachers with feedback, including suggested modifications to instruction. Well-designed assessment can have tremendous impact on students’ learning of science if conducted regularly and used by teachers to alter and improve instruction.
Teachers and Schools
Conclusion 13: To create a successful science classroom, teachers need to modify and adapt curriculum materials so as to design instruction that is appropriate for a particular group of students at a particular time. Making these kinds of modifications to achieve effective instruction requires knowledge of science, knowledge of how students learn science, and knowledge of how to plan effective instruction. Many K-8 teachers have insufficient knowledge in one or all of these areas and need ongoing support to develop it.
The demands on teachers of providing effective science instruction are immense. As no curriculum can remove teacher decision making from instruction, enacting high-quality science instruction broadly will require dramatic improvements in all three areas of teacher knowledge.
First, teachers must understand the science they teach broadly and deeply, including mastery of the four strands of proficiency we have described for student learners. This broad understanding of science is not readily supported by the typical undergraduate science courses provided for aspiring teachers. Accordingly, although increasing the undergraduate science course requirements for prospective teachers may bolster teacher knowledge in some important ways, it is unlikely to provide them with sufficient understanding of science unless the courses are redesigned.
Second, teachers need to understand the current intellectual capabilities and developmental trajectories of their students. As instruction should tap students’ existing and emergent skills and build on their conceptual knowledge base, teachers need to understand how students think, what they are capable of doing, and what they could reasonably be expected to do under supportive instructional conditions, and how to make science more accessible and relevant to them.
Third, teachers need specialized science knowledge about teaching science in order to bring their understanding of science and students’ capabilities together in well-crafted learning experiences. To plan instruction and monitor student progress, teachers need to understand how to elicit and interpret students’ understanding. They must be able to harness their understanding to inform instruction both in real time and throughout the academic year. They need to understand what students find confusing or difficult as well as what they find interesting. Furthermore, teachers need a repertoire of instructional strategies, curricular examples, and knowledge of curricular and reference materials to draw on in planning and providing instruction.
Developing these three areas of knowledge requires professional development that is both rich in science content and closely linked to teachers’ classroom practice.
Conclusion 14: Achieving science proficiency for all students will require a coherent system that aligns standards, curriculum, instruction, assessment, teacher preparation, and professional development for teachers across the K-8 years.
In effective science classrooms, curriculum, instruction, and assessment form an instructional system that is integrated. In these classrooms, students encounter a curriculum that engages them with scientific knowledge and practice in challenging and stimulating ways and flows logically and coherently across grades K-8. Current science curriculum standards have provided some focus and long-term vision for curriculum sequencing. However, they are still too numerous, loosely integrated across topics and aspects of science (e.g., inquiry practices and science concepts), and insufficiently specified to drive a cohesive instructional system. Moreover, new research on student learning suggests that there are areas in which the standards underestimate students’ capabilities to learn and do science.
A well-designed instructional system provides students with opportunities to learn science that are aligned with summative assessments. In these systems, day-to-day instructional decisions are informed by classroom-based formative and benchmarking assessment practices that provide snapshots of students’ emerging understanding.
Professional development that supports quality science instruction is ongoing, rooted in the science that teachers teach, and relevant to their classroom contexts. It provides teachers with opportunities to think and work collectively on instructional problems, supporting their efforts to tailor curriculum and instruction to classroom contexts. Professional development that supports instructional improvement rests on school- and system-level commitments that are manifest in actively involved leadership and the establishment of regular times throughout the school day for teachers’ collaboration.
Diversity and Equity in Science Education
The committee is unanimous in emphasizing the pressing need to understand the sources of inequity in science education and to identify strategies for eradicating these inequities. However, we concluded that, given the complexity of the issue and the state of the evidence base, it would be premature to formulate a set of specific findings and recommendations in this area.
The committee began its deliberations with a focus on learning and instruction at the classroom level. The evidence makes clear that all students, regardless of background, have the capabilities needed to engage with and be successful in science. Evidence from research also indicates that students from varied cultural, linguistic, and socioeconomic backgrounds bring different resources to the classroom, which must be attended to in instruction (see Conclusion 6). In fact, many studies demonstrate that it is possible for traditionally underserved children to learn science with understanding with improved instruction. Yet there is little or no agreement in the literature about the degree to which instruction should be modified for children from different backgrounds, nor what such modifications should look like. For this reason, the committee was unable to arrive at conclusions or recommendations related to instruction for diverse student populations. We did agree, however, that further research is needed to examine the effectiveness of different instructional approaches, whether these approaches are complementary or competing, and whether each approach is more effective in different instructional contexts (see the section below called “Agenda for Research and Development” for further discussion).
As the committee began to look beyond instruction to consider additional sources of achievement gaps and inequities in science education, the importance of systemic issues—such as inequities across schools in qualifications of teachers, facilities, and resources—became apparent. As we began to explore the relevant research literature, it became clear that it is difficult to tease out which systemic issues are unique to science education and which are pervasive issues that cut across all subject areas. What
is more, the complex interplay among cultural, linguistic, and socioeconomic issues is difficult to document and understand. Given the scope of this study and limited time, the committee was unable to undertake a thorough review of this body of evidence. For these reasons, we did not develop conclusions or recommendations related to the systemic issues that contribute to inequities in science education. We stress that inequities in the quality of instruction, the qualifications of teachers, resources, facilities, and time devoted to science are unacceptable and must be addressed. We identify this as a critical area for further research (see research agenda section below for further discussion). While current research does not provide sufficient guidance we urge policy makers, education leaders, and school administrators to join researchers in examining and revising policies and practices in schools and districts so that existing inequities are better understood and can be eliminated.
RECOMMENDATIONS FOR POLICY AND PRACTICE
Based on our findings and conclusions, a new view of science education is needed for K-8 schools. It should build on the new insights and reconceptualizations about how children learn science provided by the past 30 years of research in cognitive and developmental psychology and science education. These insights about learning require changes in standards, curricula, instruction, and assessment so that they are organized around the four-strand model of science learning and build the core ideas of science in a cumulative fashion across the K-8 grades. In this section, the committee lays out key steps toward realizing this new vision of science education.
Our recommendations for action are grounded in the evidence base reviewed in this report. However, in some areas the research base is not robust enough to offer a detailed, step-by-step roadmap for improving all aspects of science education. Given the urgent need for improvement and the potential power of approaches identified in emergent research, the committee focused on “best bets” for the next steps of policy, research, and practice. These best bets represent the most promising directions forward, based on the best research evidence available. They require additional documentation through continued research and careful evaluation of implementation. Through a substantial research and development effort that includes evaluation of school, district, and state initiatives, these best bets can be transformed into well-researched alternatives for policy and practice.
In framing our recommendations for policy and practice, the committee takes the perspective that science standards, curriculum, assessment, instruction, and teacher professional development should be conceived of, designed, and implemented as a coordinated system. In this view, stan-
dards and curriculum should lay out specific, coherent goals for important scientific ideas and practices that can be realized through sustained instruction over several years of K-8 schooling. Assessment should provide teachers and students with timely feedback about students’ emergent thinking that, in turn, supports teachers’ efforts to improve instruction. Teacher preparation and professional development should be focused on developing teachers’ knowledge of the science they teach, how students learn science, and specific methods and technologies that support science learning for all students.
Standards, Curricula, and Assessment: What to Teach and When
Recommendation 1: Developers of standards, curriculum, and assessment need to revise their frameworks to reflect new models of children’s thinking and take better advantage of children’s capabilities. Standards and many widely used curriculum materials fail to reflect new evidence about children’s thinking, particularly the cognitive capabilities of younger children, which are greater than previously assumed.
Recommendation 2: The committee thinks that the next generation of standards and curricula at both the national and state levels should be structured to identify a few core ideas in a discipline and elaborate how these ideas can be grown in a cumulative manner over grades K-8. Focusing on core ideas requires eliminating ideas that are less central to the development of science understanding. Selection of the core ideas should be guided by their status as foundational ideas in the disciplines of science that connect to many related scientific ideas, as well as the potential for sustained exploration at increasingly sophisticated levels across grades K-8. While existing national and state standards have been a critical first step in narrowing the focus of science in grades K-8, they do not go far enough. Future revisions to the national standards—and the subsequent interpretation of these standards at the state and local levels and by curriculum developers—need to clearly identify the knowledge and practices that are most central to the disciplines and describe how these can be developed over successive grades based on current models of children’s learning.
Recommendation 3: Developers of curricula and standards need to present science as a process of building theories and models using evidence, checking them for internal consistency and co-
herence, and testing them empirically. To this end, discussions of scientific methodology need be introduced in the context of pursuing specific questions and issues rather than as templates or invariant recipes. Similarly, the methodology students encounter in the classroom needs to reflect the range of investigatory forms in science. This requires expanding beyond a focus on experiments to incorporate examples from disciplines of science that employ observational methods or historical reconstruction.
Instruction: How to Teach
Recommendation 4: Science instruction should provide opportunities for students to engage in all four strands. This requires policy makers, education leaders, and school administrators to ensure that adequate time and resources are provided for science instruction at all grade levels for all students. They must also ensure that teachers have adequate knowledge of science content and are provided with adequate professional development.
Recommendation 5: State and local leaders in science education should provide teachers with models of classroom instruction that incorporate the four strands. These models should incorporate examples of instruction that provide opportunities for interaction in the classroom, where students carry out investigations and talk and write about their observations of phenomena, their emerging understanding of scientific ideas, and ways to test them.
Professional Development: Supporting Effective Science Instruction
We call for a dramatic departure from typical professional development practice both in scope and kind. Teachers need opportunities to deepen their knowledge of the science content of the K-8 curriculum. They also need opportunities to learn how students learn science and how to teach it. Teachers need to know how children’s understanding of core ideas in science builds across K-8, not just at a given grade or grade band. They need to learn about students’ entering conceptual ideas and ideas about science itself. They need to learn how to assess children’s developing ideas over time and how to interpret and respond (instructionally) to the results of assessment. Teachers need opportunities to teach science as an integrated body of knowledge and practice (the strands of scientific proficiency).
Teacher preparation and professional development that sensitize teachers to the capabilities of all learners and which develop teachers’ capacity to
effectively teach science to diverse student populations are urgently needed. In order to provide adequate opportunities for all students to learn science, teachers need preparation and professional development in how to respond to variation among students. They also need to know how to recognize and build on the strengths and needs that students of diverse cultural and linguistic backgrounds bring to the classroom. This should be a central feature of science teacher preparation courses and ongoing teacher professional development.
Providers of professional development should align their programs with the key conclusions and recommendations in this report. They should pay particular attention to the four strands of scientific proficiency, building on core ideas in science over long periods of time, and current research on how students learn science.
Recommendation 6: State and local school systems should ensure that all K-8 teachers experience sustained science-specific professional development in preparation and while in service. Professional development should be rooted in the science that teachers teach and should include opportunities to learn about science, about current research on how children learn science, and about how to teach science.
Recommendation 7: University-based science courses for teacher candidates and teachers’ ongoing opportunities to learn science in service should mirror the opportunities they will need to provide for their students, that is, incorporating practices in all four strands and giving sustained attention to the core ideas in the discipline. The topics of study should be aligned with central topics in the K-8 curriculum so that teachers come to appreciate the development of concepts and practices that appear across all grades.
Recommendation 8: Federal agencies that fund providers of professional development should design funding programs that require applicants to incorporate models of instruction that combine the four strands, focus on core ideas in science, and enhance teachers’ science content knowledge, knowledge of how students learn science, and knowledge of how to teach science.
AGENDA FOR RESEARCH AND DEVELOPMENT
In our synthesis of the research evidence, we have drawn on varied programs of research across multiple fields, all of which can be brought to bear on the question of how children learn science. Integrating across bod-
ies of evidence is necessary because the problem of how best to design science instruction to support a deep understanding of science is inherently interdisciplinary. In some cases the committee considered programs of research that were explicitly concerned with science learning and instruction in school settings. In other cases the research was designed to investigate fundamental questions about how children come to understand and respond to the world around them. Making sense of such a broad body of research that is often informed by different theories and different methodologies is challenging.
A critical question for continuing to advance understanding of how to support science learning and instruction in schools is how to organize programs of research so that they explicitly address problems of educational practice in schools while advancing fundamental understanding of children’s learning in science. Two key elements of such a program that need to be thought through are (1) What is the nature of the teams of individuals who should be brought together to conduct this work? and (2) How can some common intellectual ground be developed so that a dialogue can begin across the varied research traditions?
The committee agrees that there is a glaring lack of an infrastructure for research, development, and implementation in science education that is informed by research on fundamental aspects of learning and teaching but takes up problems and questions that are grounded in the realities of practice. This research and development effort must be closely tied to schools and classrooms. Research and development partnerships must include teachers, administrators, curriculum developers, providers of professional development, and district- and state-level supervisors. Funding streams must support studies at various levels, including design and development work to identify promising approaches, small-scale testing of initial concepts under controlled conditions to establish viability, classroom-based research in a few classrooms or schools, replication to explore the implications of varying conditions, longitudinal studies, and finally implementation and evaluation on a large scale.
Critical Areas for Research and Development
Learning Across the Four Strands
The four-strand framework represents a departure from the way research on science learning has been organized in the past. Researchers have tended to focus on either domain-specific learning or domain-general reasoning related to Strands 1 and 2 in our model. More recent work has begun to look at Strands 3 and 4, and further research is needed in these areas. For example, an area that needs increasing attention is related to students’ un-
derstanding of how scientific knowledge is constructed and how they come to understand and negotiate different knowledge communities. That is, how do children start to navigate the terrain of knowledge around them? How do they know who is credible and who is not? How do they determine who is a trustworthy source? Also of interest is a related set of questions about students’ understanding of the status of their own knowledge, such as, How do you know when you don’t know?
Much more research is also needed to further elaborate the interconnections between the four strands—for example, studies of the interplay between domain-specific and domain-general knowledge over the course of development. Understanding interconnections between the strands and how instruction might better leverage these interconnections is of particular interest for informing instructional models based on the four strands.
Identifying Core Ideas and Developing Learning Progressions
Developing learning progressions to structure science standards, curricula, instruction, and assessment is a promising direction for science education, but an extensive research and development effort is needed before learning progressions are well established and tested. A major first step is to identify the most generative and powerful core ideas for students’ science learning (i.e., those that have broad reach across science disciplines and provide the best leverage for students’ future study of science) through a cross-disciplinary research program. From these core ideas a series of learning progressions can be developed and tested. Research and development necessary to establish the empirical basis for learning progressions across the domains of science will need to include multiple phases, including focused studies of children’s learning under controlled conditions, small-scale instructional interventions, classroom-based studies in a variety of contexts, and longitudinal studies.
Longitudinal studies over multiple ages are particularly important. In a given domain and across domains, a better understanding is needed of continuities and discontinuities in students’ understanding across grades K-8. That is, what are the legacies of early development, and what is new and different as children develop and encounter different experiences both inside and outside the classroom? What are the mechanisms behind the changes? Finally, for a given set of related concepts, the research should examine the trade-offs of different learning sequences or instructional approaches as well as the instructional support needed to help students move through the progression.
Curriculum and Instruction
Studies of instruction and the links to student learning are needed to develop instructional models that integrate the four strands. Research is also
needed to develop a better understanding of whether and how instruction should change with children’s development. Clear depictions of scientific practice across K-8 and variations among particular practices across consecutive years of instruction (e.g., younger versus older children’s argumentation) should be developed. One mechanism for deepening the understanding of effective instruction is through replication of classroom-based research on instruction (e.g., design studies). Two additional areas of particular need are (1) the development of tools to help teachers diagnose students’ understanding and cue-productive instructional options for teachers to advance it and (2) the characteristics of instruction that best serve diverse student populations.
Research on curriculum materials is also a critical area. Such studies should systematically analyze the effects on learning of variation in conditions, such as student populations, school settings, teacher knowledge, and forms of professional development, as well as the dimensions on which curricula vary (i.e., comparing curriculum focused on content knowledge, on contextualized science problems, on modeling). Longitudinal studies of different curriculum approaches under varied conditions would be particularly useful.
Professional Development and Teacher Learning
A substantial commitment is needed to empirical research on the practices of building expertise in science teaching. These include using science specialist teachers (in K-5), mentoring, teacher work groups, instructional materials designed to support teacher learning, and long-term professional development. It is important to understand how local circumstances enable or limit the effect of these models. This research needs to establish an empirical relationship between professional development and student learning.
Evaluation and Scale-Up
Evaluation of current and emerging instructional practices and curriculum materials is a critical part of the research and development cycle. We stress that it is preferable to develop a substantial research base documenting the effectiveness of a particular approach before it is taken to scale. Often, however, this is either not done in practice, or it is not feasible given the pressing needs of schools and districts for immediate solutions. In such cases, systematic evaluation efforts should be tied to large-scale implementation of instructional practices and curriculum materials that are widely implemented before adequate small-scale testing is complete. Such evaluations must be carried out in partnerships with school systems and states. Capitalizing on the increased availability of student data due to the reporting
demands of the No Child Left Behind Act of 2001 (H.R. 1) might be a useful way to investigate the impact of particular approaches at scale.
Diversity and Equity
Research on supporting science learning for culturally, linguistically, and socioeconomically diverse students is an area of critical need. This includes research on instruction, curriculum assessment, and professional development. For example, alternative instructional approaches have been proposed to promote science learning with nonmainstream students with promising results. Further research is needed to examine the effectiveness of each approach, whether these approaches are complementary or competing, and whether each approach is more effective in different instructional contexts. In addition, research on curriculum models that incorporate effective approaches to instruction for diverse students is necessary.
More work is also needed on understanding systemic factors involved in creating inequitable learning opportunities in science. For example, differences across schools in teacher qualifications, resources devoted to science, and time for science instruction should be explored. The interactions of culture, ethnicity, language, and socioeconomic status in shaping students’ opportunities to learn science are also important areas for further research.
We live in a time when science is a ubiquitous part of civic and political life. The pressing issues of today—global warming, pandemics, alternative fuels, use of biometric information to fight terrorism—require a scientifically informed citizenry as never before in the nation’s history. Calls for a better prepared scientific and technical workforce have become more urgent in the context of increasing globalization and fears that a diminished capacity for innovation will make the United States less competitive in a global market. With the convergence of these issues, the quality of science education in this country takes on tremendous importance. Yet students’ performance in science is disappointing, and recent improvement efforts have proven insufficient. What is more, non-Asian minority and disadvantaged students are consistently among the bottom-performing groups, trailing economically advantaged and white students. Such achievement gaps are unacceptable in view of the increasing diversity of the American population, the reality that science permeates society at all levels, and overwhelming evidence that children from all backgrounds have the capacity to become proficient in science.
To improve science education in the United States, changes are urgently needed throughout the system. The evidence reviewed in this report provides a compelling framework for how science education can be reshaped to take account of research on how best to support children’s science learning. Admittedly, further research is needed, especially to advance the strands
framework, elaborate the learning progressions, increase understanding of effective approaches to instructional design, and determine how best to support teachers. Nevertheless, the research base reveals that current approaches are inadequate, and it provides a roadmap for moving forward. Beginning with what is now known about how children learn science, the direction for teaching and for the education of teachers is clear.