AAAS - Project 2061 - Designing Systems to Support Learning Science with Understanding for All

Designing Systems to Support Learning Science with Understanding for All: Developing Dialogues among Researchers, Reformers, and Developers

Andy Anderson

January, 2001

Science education at both the K-12 and postsecondary levels is currently being reshaped by two long-term trends. One of these is the standards movement, which seeks to build consensus about goals and methods for science teaching and encourage their large-scale adoption. These standards are instantiated in authoritative documents such as the National Science Education Standards (NSES) and Benchmarks for Science Literacy, as well as other standards at the state and local levels, teacher proficiency standards, and so forth. The second trend is the development of information technologies which provide new resources for science teaching and learning and new opportunities for economies of scale.

It is clear that these trends are affecting teaching and learning in science classes. Teachers and professors are changing the content of their courses and their teaching methods in response to standards. They are using information technologies to organize their courses, to give students access to resources, and to create new kinds of "virtual" courses that do not require students and teachers to come together in the same room at the same time.

It is less clear whether these changes will be improvements. Will these developments promote student understanding, or will they provide new support systems for uninspired teaching leading to shallow learning? Will they give students who are currently marginalized new access to the benefits of scientific literacy, or will they provide new advantages to already privileged students? Will they make science teaching richer in resources for sense making, or will they promote "efficient" ways to deliver conventional content to large numbers of students?

I believe that the answers to these questions will depend in part on the degree to which work in these trends can be informed by dialogue with research on science teaching and learning. As I discuss below, the record so far is mixed at best. The development and implementation of the standards has been accompanied by some sincere attempts to consult and use research, especially conceptual change research, but much of the process has involved political consensus-building in which research and researchers were marginalized. Some development work using information technologies has used and contributed to research on science teaching and learning, but many projects have relied mostly on the instincts and experience of their developers, It is not surprising that dialogue among these three movements-the standards movement, development of information technologies, and research on science teaching and learning-is difficult. Researchers, reformers, and developers often come from very different backgrounds and use different languages to describe and analyze their work.

I want to make a case, though, that the rewards of such a dialogue are worth the challenges. The purpose of this paper is to outline two broad sets of issues around which the current dialogue among these three movements might productively be deepened and enriched, and to suggest contexts in which that enhanced dialogue might occur. The paper is written primarily from my perspective as a researcher who has some awareness of and involvement in the other two movements. My suggestions about the basis for an enhanced dialogue have three parts:

Goals: Participants in all three movements share commitment to a common goal of learning science with understanding for all. Research in science studies and science education suggests some basic ideas about science understanding as sense making about patterns in experience. These ideas could be used to enrich and deepen our dialogue around this common goal.
Challenges: Some fundamental challenges are inevitable for person or system that attempts to help large numbers of students learn science with understanding. Science education researchers have explored four challenges that "come with the territory" of teaching science for understanding for all. These common challenges could provide a basis for dialogue among people taking different approaches to teaching science for understanding for all.
Opportunities for dialogue. Participants in the three movements cannot stop their current work in order to talk to one another. I suggest some ways in which dialogue around our common goals and challenges could be built into research and development activities in which we are already engaged.

Goals: Learning Science with Understanding for All

Almost everyone involved in science education research and development claims "learning with understanding for all" as our basic goal. However, our apparent consensus is based in part on the vagueness of that phrase. Because we can all attach our own implicit meanings to the phrase, we can all agree that this is what we want. I would like to suggest some ways in which we might make the meaning of this phrase more precise and connect it to the research literature in science studies and science learning. In particular, I want to argue that we could take "understanding" to mean something like sense-making about patterns in experience. For me, this phrase captures the essence both of what scientists do and of what science educators want learners to do. I will use this section to elaborate on this idea.

The task of science

I would like to start my case with a quote that David Hawkins uses from Niels Bohr:

The task of science is both to extend our experience and reduce it to order, and this task represents various aspects, inseparably connected with each other. Only by experience itself do we come to recognize those laws which grant us a comprehensive view of the diversity of phenomena. As our knowledge becomes wider we must always be prepared, therefore, to expect alterations in the points of view best suited for the ordering of our experience.

There are many things I like about this quote, both as a description of "the task of science" and as a description of the task of science education. Let me start with one. In this quote Bohr suggests an epistemological position that is intermediate between modern empiricism and postmodern relativism. On the one hand, Bohr does not suggest that through our experience we can ever fully understand the world as it is. There will always be parts of the world that are beyond the range of our experience, differences in our ways of experiencing the world, and differences in our ways of reducing experience to order. On the other hand, Bohr suggests a way of judging the quality of theories or world views that moves beyond relativism. As Bazerman suggests, theories that encompass a more extensive range of experience or more thorough and thoughtful attempts to reduce that experience to order deserve special respect.

Western scientific subcultures have been especially successful in extending the collective experiences of their members and reducing those experiences to order. This success has been based in part on a set of epistemological distinctions that are woven into the practice of Western science. In particular, the program of Western science involves coordinating knowledge claims of three different types:

Experience in the material world (data). We think of the material world as consisting of systems and phenomena that are the basis for our experience, but we can know these systems and phenomena only through our interactions with them--through our experience in the material world. The success of Western science has been based in part on a decision not to try to account for all experiences in the material world, but instead to concentrate on experiences that have been verified, reproduced, described or measured precisely, recorded, and shared--in other words, data. Thus data are created from selected and refined experience. Data are created, not merely collected.
Patterns in experience. Scientific facts, laws, and generalizations are statements about patterns that scientists see in their data. Thus pattern finding is an essential scientific practice, a key step in "reducing our experience to order." These patterns in experience are the essential links between data and theories.
Explanations of patterns in experience. Scientific models and theories are metaphors, or stories, or symbolic systems that are designed to explain patterns in experience. Although they can be used to explain individual phenomena, their power lies in the fact that they provide parsimonious accounts of broad patterns that encompass many different systems or phenomena.

In pursuing the task of science--both extending our experience and reducing it to order--members of scientific communities routinely recognize and use the epistemological distinctions described above. That is, they share common meanings for key terms such as data, laws, models, and theories, and they use those terms to organize their professional practices. Thus these distinctions have been essential to the success of Western science.

There are many important similarities between this view of science and the science education consensus views as presented in documents such as Science for All Americans and the National Science Education Standards. I think, though, that there are three ways in which this view differs in emphasis or approach from the consensus view. This view presents data creation as active selection and refinement of experience; it emphasizes the importance of patterns in experience; and it presents a unified view of scientific practice and scientific knowledge.

Data creation as active selection and refinement of experience. The consensus view emphasizes the importance of data and the complex relationship between data and theories. It is largely silent, however, on the complex epistemology and creative practice involved in "data collection." In contrast, the science studies literature reveals the process of creating data to be a complex enterprise that is intensely controversial in scientific communities. Three core problems are especially salient. First, scientific communities must separate data from noise. That is, they must decide which experiences merit their attention and which experiences they will dismiss as distortions of the "true nature" of the material world. Standards such as reproducibility and precision play an important role in this process, but ultimately decisions about which experiences count as data are negotiated within scientific communities. The second core problem concerns techniques for data creation. The most precise and reproducible data rely on methodologies and instruments that are designed to provide new experiences that extend our senses and create new phenomena for observation. A third core problem is data representation. Scientists must invent ways of representing and communicating about their experiences that allow others to share in those experiences and make patterns apparent. This is an active, creative, process, not just an exercise in recording and displaying measurements.

The importance of patterns in experience. The consensus view emphasizes the distinction between data and theories-these are different types of knowledge claims that are created and used differently by scientific communities. However, the consensus view does not emphasize the importance of pattern-finding practices and their products (e.g., laws, generalizations, data displays) as ways of "bridging the gap" between data and theories. I think that this is a mistake. Pattern finding practices lie at the core of the scientific enterprise, patterns in data are the essential elements that create coherence and parsimony in scientific knowledge claims; and the applications of scientific knowledge rely on our ability to find and use the appropriate patterns.

A unified view of scientific practice and knowledge. Although the consensus view sees science content, scientific inquiry, and the nature of science as intimately related-all part of the same big picture-these domains of scientific understanding tend to be represented separately. They draw on different bodies of scholarly literature and use different theoretical frameworks. The view above suggests a way in which we could talk about these three domains in a coherent and parsimonious way:

Scientific inquiry consists of the set of practices by which scientific communities create shared bodies of experiences with the material world, patterns in those experiences, and explanations for those patterns. In this view, the "scientific method" is a formalized form of argument. A research report that includes questions, hypotheses, methods, results, and conclusions must combine experiences, patterns, and explanations in a coherent argument that can be understood and discussed by other members of the scientific community.
Scientific content consists of the results of the cumulative practices of scientific inquiry-a body of experiences, patterns, and explanations that have been accepted by consensus of scientific communities.
The nature of science is our label for our attempts to understand the history, nature, and limitations of the cumulative efforts of scientific communities. What kinds of experiences have scientific communities paid attention to and what kinds of experiences have they ignored? What are the historical origins of our present practices and bodies of knowledge? How can we compare the experiences, patterns, and explanations of Western scientific communities with those of other communities?

One final note about the task of science. Though the discussion above portrays scientific practice and knowledge as complex, the goal of sense making about patterns of experience with the material world is the unifying core of those practices. Scientific communities engage in collective sense making by creating coherent systems of experiences, patterns, and explanations.

The task of science education

We include science in the school curriculum because of the success of the Western scientific enterprise. Scientific communities have devised ways of accumulating vast stores of experience with the material world, of selecting and refining those experiences into precise and reproducible data, and of recording, sharing, and preserving those data. They have been successful in finding patterns in those data and in developing theories to explain those patterns. We want our students both to see and use patterns in their own experiences with the material world and to benefit from the cultural heritage accumulated by their forebears.

Thus Bohr has described the task of science education as well as the task of science. Science educators should help learners to extend their own experiences-both personal and vicarious-and reduce them to order. For students, as for scientists, learning with understanding requires sense making about patterns in experience. I would like to suggest two complementary definitions of understanding, one focusing on interactions with the material world, the other focusing on social interactions.

Understanding as sense-making in the material world. If we look at the NSES content standards or the Project 2061 Benchmarks, we find many statements about scientific laws or theories. What does it mean to "understand" these statements? One kind of definition would focus on understanding as a way of interacting with the systems and phenomena of the material world. We could say that understanding involves an ability to coordinate experiences in the material world, patterns in those experiences, and explanations of the patterns.

This definition implies that learners would understand a particular theory, for example, when (a) they have access to the relevant experiences in the material world, (b) they can see patterns in those experiences, and (c) they can use the theory to explain the experiences and patterns. In extending their experience they need both to interact personally with systems and phenomena-to create their own data-and to make use of data created by others. Those data created by others need to be, in Paul Cobb's words, "experientially real" to the students. Similarly, their ways of reducing experience to order need to be both personally meaningful to them and consistent with canonical scientific laws and theories.

Understanding as participation in sense-making communities. A complementary definition of scientific understanding focuses on learners' capacity to participate in communities of practice that are engaged in sense making about the material world. Understanding in this sense has canonical, participatory, and personal dimensions.

Canonical understanding involves being able to discuss experiences, patterns, and theories in terms that members of scientific communities would recognize and approve.
Participatory understanding involves being able to play a productive role in a community of practice that is collectively making sense of experience in the material world. Some individuals who have serious gaps in their canonical understanding may participate quite successfully in collective inquiry. Conversely, some individuals who have excellent canonical understanding may not be able to communicate with other members of a group or contribute to its work.
Personal understanding involves making sense in ways that are personally meaningful. Some students who can reproduce canonical explanations may not find them personally satisfying. Conversely, some students may be quite satisfied with explanations or sense-making styles that are not recognized by scientific communities. Other communities may recognize those explanations or sense-making styles as valid.

Understanding for all. For me, a goal of "understanding for all" entails obligations in each of the senses discussed above. Teaching for understanding obligates us to extend the depth and breadth of learners' experiences in the material world and their resources for making sense of those experiences. Sense making is a collective as well as a personal enterprise, so we are also obligated to help learners participate in communities of practice that produce and use data, patterns, and explanations. We can have legitimate differences of opinion about the relative importance of these different criteria for understanding, but we must recognize that all of them are important for our students.

Dialogue around scientific understanding for all

For me, the import of the discussion above is that "scientific understanding for all" has a simple core meaning around which many complexities and ambiguities accumulate. The simple core meaning is this: Scientific understanding is a product of personal and a social sense making in which we "extend our experience and reduce it to order." That is we (a) create data out of our experiences with the material world, (b) find patterns on those data, and (c) explain those patterns. When we say we want understanding for all, we declare our intention to allow all students to participate in this personal and social sense making and to gain access to the body of data, patterns, and explanations created by scientific communities. The complexities arise when we begin to study the various ways in which scientific communities (and other communities of practice) have extended their experience and reduced it to order, as well as the various ways in which individual students or classroom communities might themselves make sense of the world. There will always be inescapable value judgments as we decide which practices count as "real" understanding.

Researchers, reformers, and developers currently use many different frameworks and vocabularies for discussing scientific understanding. These differences tend to obscure both the common core meanings and the surrounding complexities. The lack of a deeper and more precise vocabulary makes it difficult either to find common ground or to understand our differences. This is my attempt to suggest some ways of understanding "understanding" that might help us to achieve a deeper consensus about our areas of agreement and to understand better how we disagree.

Challenges: Teaching for Understanding in an Information Age

Let's think of the goal of scientific understanding for all as a challenge to classroom systems or classroom communities. Teachers are the leaders of these classroom communities. Supported by textbooks, laboratory materials, and other resources, teachers need to help their students extend their experiences in the material world and reduce them to order. That is, teachers need to help their students construct for themselves coherent systems of experiences, patterns, and explanations.

Research on science teaching and learning has documented many examples of successes and (mostly) failures of classroom communities as teachers have tried to lead their students toward understanding for all. In many cases, the failures can be attributed to lack of appropriate goals or of resources to achieve those goals. If the teacher doesn't understand a relevant theory and the textbook doesn't explain it adequately, for example, then students aren't likely to make sense of the theory. Similarly, if students access to experientially real data (real or vicarious), then they are unlikely to see the related theories as tools for sense making-ways of explaining patterns in experience.

New standards and information technologies could help us to address these problems. Standards could help us agree on goals that are reasonable and appropriate. Information technologies could be used to make new resources available to classroom communities. Students could have access to new tools for creating data and finding patterns in their own experiences. They could gain access to vicarious experiences in many different forms. They could construct and compare their own explanations or find explanations appropriate to their own needs as learners. In theory, at least, information technologies could greatly reduce the resource limitations of class communities.

So what might we need to DO with these new resources? What enduring challenges do classroom communities face in helping their members to learn science with understanding? I read the research on science teaching and learning as suggesting four fundamental challenges arising from the nature and diversity of science learners:

Creating appropriate curricula,
Developing sufficient learning activities and resources,
Accommodating different styles of sense-making, and
Cultivating students' interest and effort.

1. Creating appropriate curricula

What's the right content to teach? Teachers have to select or create goals and learning experiences that are scientifically important, yet meaningful to the students in their classes. What are the opportunities for sense making about patterns in experience that will work best for a particular group of students? What is the proper balance between personal sense making and mastery of canonical practices and knowledge claims? Conceptual change and sociocultural research reveal that there are often serious mismatches between the sense-making styles and capacities of students in science classes and the content that is taught in those classes. These mismatches often lead to situations where students stop trying to make sense of the content that they are being taught. They resort to procedural display or stop trying entirely.

For example, one very common problem is that theories are taught to students who have little or no exposure to the experiences and patterns that those theories explain. Teaching the structure of the atom is one example. Students commonly are required to memorize quantum numbers and the structure of electron shells in courses where they will have little or no access to the patterns that these theories explain, such as the three-dimensional shapes of molecules or patterns of emission lines in rare gases. Even patterns of valence and electronegativity in the Periodic Table (which could be explained with less complex theories) are far removed from any phenomena that are experientially real to most students.

Even experiences that are accessible to most students are not necessarily ones that they have refined into data and noticed patterns in. For example, few students can describe the apparent path of the sun across the sky, or patterns they have noticed in floating and sinking of liquids in one another, or the growth of plants and animals in their back yards. Not all students have back yards, of course. Social and cultural diversity make the task of deciding what experiences, patterns, and theories are appropriate for a particular class particularly challenging.

Documents such as the National Science Education Standards and Benchmarks for Science Literacy contain consensus suggestions about which experiences and theories are appropriate for students of different ages. These documents, though, are based mostly on the experiences of their writers and consultants, with a little bit of help from conceptual change research. A broader look at the research could lead to deeper discussions of how we might select and present content for particular groups of students.

2. Developing sufficient learning activities and resources

How much teaching about a topic is enough? Developing coherent systems of experiences, patterns, and explanations is a slow and difficult process for most students. They may need to extend their experiences (either personally or vicariously) and create data from those experiences. They may need to learn new ways of recording or displaying data. They may need to see or be introduced to patterns in data that they have never noticed before. They may need, as Bohr suggests, to undergo "alterations in the points of view best suited for the ordering of [their] experience."

The research literature documents numerous cases in which teachers and textbooks have made poor choices about "breadth vs. depth" issues, almost always in the direction of going on to the next topic too soon, before students have made sense the topic that they are studying. In many cases, this may be partly a resource limitation: If the textbook is your only resource and you have reached the end of the chapter, then what is there to do except go on to the next chapter? The availability of new resources does not, however, modifies rather than eliminates the problem. Teachers must still decide when a community has reached the point of diminishing returns on one topic and it is time to move on.

Even more important is the design challenge of coming up with the right combinations of learning activities to support students' sense making. What kinds of experiences and activities are necessary, and in what order? There are numerous attempts in the literature to formulate generally applicable answers to this question, such as Posner, Strike, Hewson, and Gerzog's conditions for conceptual change or various versions of the learning cycle. All of them, however, have heuristic value at best. They suggest categories of experiences or activities that might help students' understanding; they do not provide prescriptions that teachers or designers of classroom support systems can follow. Thus the challenge of creating systems that are sufficient to support students' sense making is an enduring one for teachers and curriculum developers. The research literature can provide important guidance for this empirical process.

3. Accommodating different styles of sense-making

What about student diversity? Current curricula that present science content as sequences facts and concepts to be learned or problem-solving skills to be mastered tend to divide students into three broad groups. A small number of students are able to incorporate the symbol manipulation and facts into their own sense-making strategies. Because the facts and symbols make sense to the students, they are able to learn them rapidly and remember them easily. They can proceed rapidly through the established curriculum. A second group of mostly middle-class students is not particularly successful in making sense of the symbols and facts, but they are expected by their parents and peers to be successful in school, and they have a set of strategies for memorization and symbol manipulation (especially in arithmetic) that they learned both at home and in school. Their parents provide or find extra help if they fall too far behind. These students generally make reasonably good grades through "procedural display," even without understanding the symbols and facts in any depth. A third group of students, including many lower-class, ESL, and special education students, lacks the resources or the incentives to engage in successful procedural display. These students are likely to show active or passive resistance and alienation.

Given this skills-based curriculum and this range of student responses, there seems to be little reason to put these three groups together in the same classroom. If they are tracked, the students in the first group can proceed much more rapidly through the curriculum. The students in the second group can learn procedural display much more effectively if the resistant and alienated students are removed. For the students in the third group, dealing with their resistance and alienation becomes a more salient problem than teaching science and math content. Thus tracking is a sensible response to the problems created by a curriculum that focuses on symbol manipulation and fact learning.

A view of science as sense making about patterns in experience suggests other possibilities. The science studies literature suggests that written records such as scientific journals and textbooks are formalized records of much more complex and culturally embedded sense-making processes. Scientific communities use symbols and facts as they search for patterns in the phenomena of the material world and use them in the service of real-world problem solving. The reasoning and language involved in these sense-making and problem-solving processes are often personal, informal, and imaginative.

Researchers differ in their assessment of this different view of the scientific enterprise for science teaching and learning. Some researchers, such as Yerrick, Ogbu, and many feminists point to the deep cultural divide and differences in power between scientific communities and marginalized students. They suggest that new standards or resources may do little to alter enduring patterns of resistance and alienation in the absence of larger systemic changes.

Other researchers, such as Warren and Rosebery or Lehrer and Schauble, take a more optimistic view. They suggest that this view of the curriculum reveals "generative continuities" between the language and reasoning of children, including children who are poorly prepared for procedural display, and the language of reasoning of scientists and mathematicians. Warren points specifically to "embodied imagining" and informal reasoning as examples of those generative continuities. These researchers include the school curriculum, including the curriculum in the standards, as "the neck of the hourglass"-a narrow and restricted view of science that cuts off the generative continuities between students' and scientists' reasoning. Thus research could help us to discover generative continuities between students' and scientists' strategies and help students to use them in the service of their collective and individual sense-making efforts.

4. Cultivating students' interest and effort

Why should students bother to do the hard work of learning with understanding? While it is true that students are naturally curious about the world around them, many students are not naturally inclined to engage in the kinds of sustained effort necessary to develop coherent systems of experiences, patterns, and explanations. Exciting experiences or activities are good for catching students' attention, but student motivation to engage in the work of sense making must be cultivated and nurtured.

Research on teaching and learning tells us that the issues discussed above affect students' motivation and interest. Students generally don't keep trying to make sense of science if they are being taught content that is inappropriate for them, or if the learning activities in the classroom are culturally inappropriate or insufficient to support understanding. Not surprisingly students respond better when the curricula are developmentally and culturally appropriate and the learning resources and activities are sufficient to support their sense-making efforts.

It is also clear that students are flexible in their curiosity and their willingness to work. They can become interested in seemingly arcane issues or they can find "motivating" activities dull and boring. They can respond to the enthusiasm of other members of their learning communities. Their interests in pattern finding and sense making can be cultivated over time, as can their abilities to find interesting questions in their experiences with the material world. Feminist and sociocultural researchers point out that students' interests and efforts are closely tied to their feelings about identity and empowerment. They put their efforts into activities that are consistent with the norms of their cultural groups and their images of themselves.

Dialogue around the challenges of teaching for understanding

I read the current research on science teaching and learning as documenting the importance of these four design challenges for any teacher or developer. For better or worse, they will be present whenever we try to help students learn science with understanding. The research documents a few "existence proofs" of sophisticated and well-supported teachers who meet all four of these challenges well. Mostly, though, it documents how difficult these challenges can be and how rarely our current systems meet them well.

Our current research and development efforts tend to emphasize some of these challenges at the expense of others. The standards, for example, focus primarily on the first two challenges while paying much less attention to the third and fourth. Textbooks and support systems using information technologies tend to address some of the challenges while leaving others to the craft of the teacher. The challenges are often addressed in ways that are shown by current research to be predictably inadequate. At best, though, the research provides spotty and inadequate support for developers who try to use research to meet these challenges. There are many opportunities for dialogue around these challenges to enhance our research and development efforts.

Opportunities for dialogue around goals and challenges

I have suggested two issues around which researchers, reformers, and developers might be able to develop a deeper and more productive dialogue. The first of these concerns the nature of scientific understanding. It might be possible to build a consensus around the idea that we want to engage all students in sense-making about patterns in experience. This means that, individually and collectively, students need to develop coordinated systems of experiences with the material world (data), patterns in those experiences (laws, generalizations, data displays), and explanations for those patterns (theories, models).

The second issue around which researchers, reformers, and developers might develop a deeper dialogue concerns the enduring challenges of teaching for understanding. The current research in science education suggests to me that all teachers and developers face at least four fundamental challenges. They must (a) create curricula that are appropriate for their students needs and interests, (b) develop activities and resources sufficient to support students' learning with understanding, (c) accommodate different styles of sense-making, and (d) cultivate students' interests and efforts.

It might be worth bringing researchers, reformers, and developers together for a separate conference around these shared goals and challenges. Such a conference would be worthwhile, though, only if it stimulated a dialogue that continued through the daily work of the people involved. For example, here are some ways in which discussion around these goals and challenges might inform the activities of research, reform, and development.

Analysis of science teaching and learning. These goals and challenges could provide categories around which lessons or units could be analyzed.
Evaluation of systems for teaching and learning. Systems that include teachers, textbooks, laboratories, and information technologies could be assessed on how they address these goals and challenges.
Design criteria for course development. These goals and challenges could be used to suggest design criteria for developing K-12 or postsecondary courses, including courses that rely on information technologies.
Revision of science education standards. These goals and challenges, and the research on which they are based, could be used to inform the revision of national, state, or local standards.
Science teacher education. Science teacher education programs could try to prepare teachers to engage these goals and challenges through their own knowledge and skills and the appropriate use of supporting resources.

The list above is not complete, or even very thoughtful. I hope that the underlying idea, though, is clear. Our reform and development efforts will be stronger in the long run if they are supported by a continuing dialogue with science education research. We may be able to organize this dialogue around our shared goal of learning with understanding for all and around the shared challenges that all reformers and developers must address in order to meet this goal.