[Presented at CSCL '99, December, 1999, Stanford, CA]

ChemSense: Promoting representational competence and collaborative inquiry in chemistry

Patricia Schank, Elaine Coleman, Robert Kozma

SRI International, Center for Technology in Learning

Brian Coppola

University of Michigan, Department of Chemistry

Abstract: ChemSense is a learning environment to promote representational competence and epistemological thinking in high school and undergraduate chemistry. In collaboration with students and teachers in local (California) high schools and the University of Michigan, we are developing an environment and curricular framework to help students use various representations to build models of chemical systems, and use these models to discuss and explain the chemical phenomena they represent. Our curricular framework uses scaffolded collaborative investigations as an organizing theme. Using an iterative, participatory design approach, results of students' use of ChemSense as a learning tool will be used to reengineer ChemSense so that by the end of the three-year project, we will have a set of robust computer-based tools for learning chemistry. From a research perspective, we will characterize: (a) which representations were most helpful for students to learn chemistry, (b) the relationships between representational competence, epistemological beliefs, and chemical understanding, and (c) students’ ability to transfer learning from the computer environment to other learning situations. Future work will focus on extended support for conversation between students, teachers, and mentors around student "portfolios", and multiuser facilities to support small-group collaboration.
Keywords: chemistry, representations, epistemology, inquiry, collaborative investigations
Technical requirements: Poster area and table (for demo using our own laptop)


Although chemistry plays an important role in the science curriculum, students typically leave high school with profound misunderstandings about the nature of matter, chemical processes, and chemical systems (e.g., Gabel & Bunce, 1994, Krajcik, 1991). They do not see chemistry as a process of inquiry that uses concepts and principles to explain real world phenomena. Nor do they use representations and models to think of chemical phenomena as systems of interacting molecular entities. These difficulties influence the success of students in science and their attitude about taking future science courses.

The ChemSense project attempts to address this problem by focusing on three critical and interrelated issues: chemical understanding, students' epistemological beliefs and actions, and a set of cognitive and social skills that we refer to as "representational competence" (Kozma & Russell, 1997). These skills include abilities to (a) see representations as corresponding in some way to ideas that explain a phenomena; (b) link, transform, and move fluidly among chemical representations such as structural diagrams, graphs, formulae, simulations; (c) create and use representations to think, act, and communicate, and (d) comprehend and evaluate their boundaries and limitations. We also seek to give students direct experience with methods and processes of scientific inquiry, Consequently, we are using scaffolded collaborative inquiry as an organizing theme for ChemSense to help students come to think of science as an organic process ("science in the making", Latour, 1987), reason in ways similar to those of scientists, involve them in collaboration and the social construction of knowledge, and construct explanations using multiple representations of key chemical constructs. Specifically, we are:

1. Studying the relationships between high school and college students' understanding of chemistry and their use of various representations that characterize chemical phenomena, and the relationships between epistemological thinking and students' understanding of chemistry.

2. Developing a learning environment and curricular framework, in the context of authentic lab activities and standards-based curricula and in collaboration with teachers and students. The environment, called "ChemSense," both supports and benefits from our research activities. ChemSense will allow students and teachers to enter data from their laboratory experiments which will, in turn, drive computational models of chemical systems represented by dynamic graphs and molecular animations. Students will draw on a library of representations of various sorts (formulas, structural diagrams, digital photos and movies) to express their understanding of this system, as well as construct representations of their own. Students will be able to link these representations to the model and annotate, predict, justify, explain, etc., the behavior of the model and their laboratory observations to others. The curricular framework and sample activities illustrate the range of capabilities of the environment.

3. Evaluating the educational effects of our learning environment on students' understanding of chemistry, and their ability to "transfer" their knowledge and skill of to solve other problems.

In our poster, we will present results from studies currently underway to examine the discourse of pairs of students as they reason about chemical phenomena using various chemical representations. We will also present screenshots from ChemSense activities and components currently under development, and outline our curricular framework and themes. We are creating new partnerships to co-develop and share components, research, and evaluation, and will explore potential collaborations with CSCL conference attendees.

Curricular framework and themes

We provide not only the environment and tools but also a curricular framework that will aid teachers in the development of ChemSense modules. We will use these the tools, environment, and curricular framework to develop example modules that can guide and (hopefully) inspire the work of others who choose to develop ChemSense modules.

We believe that the chemistry content of any module designed with ChemSense can be characterized by one or more of five molecular-level themes: connectivity (e.g., changes in molecular structure due to chemical reactions), shape (e.g., changes in spatial relationships that accompany chemical change), state (e.g., energy relationships and affects of heat and light), aggregation (e.g., why some things mix and others don't), and concentration (e.g., the affect on collisions and reactions). These themes can be used whether the topic of the module is a traditional one, such as "Solubility," or a novel one, such as "Making Soap." Our intent is that by using one or more of these themes in a recurrent way across modules, teachers and students will come to understand and explain perceptual phenomena in terms of underlying chemical structures and processes.

While even young students can engage in investigation, they usually require some assistance or scaffolding in the form of a standard procedure, plan, or model to guide their work. We draw upon a procedure developed by Krajcik, Blumenfeld, Marx, Bass, and Fredricks (1998) as a standard format across all of our sample ChemSense activities. This procedure can be summarize as: Ask questions, design investigations and plan procedures, construct apparatus and carry out investigations, analyze data and draw conclusions, and present findings. The components of this procedure (what Krajcik et al., 1998, call the "investigation web") address the key reasons for using collaborative investigation as an organizing theme: it engages students in science in the making, helps them reason like scientists, involves them in collaboration and social construction of knowledge, and requires them to use representations in various ways.

The ChemSense environment

Exemplary learning environments allow students to quickly build and test models of complex systems (Jackson, Stratford, Krajcik, & Soloway, in press), view and manipulate processes at various levels (GenScope, genscope.concord.org), and explore their understanding via dynamic simulations and share their interpretations with peers (Roschelle, Kaput, & Stroup, in press). Compared with other sciences, few noteworthy chemistry learning environments exist. For example, ActivChemistry (www.salamanderinteractive.com) lets students simulate chemical processes, but mainly at the level of replicating surface characteristics of lab experiments. Using ChemViz, students can generate molecular images, but have little access to the dynamics of the system, leading to weak or modest impact on learning (Moran et al., 1997). Environments that scaffold explanation and argumentation, like CSILE (Scardamalia, Bereiter, & Lamon, 1994); KIE (Slotta & Linn, forthcoming), Belvedere (Suthers, Toth, & Weiner, 1997), and Convince Me (Schank, 1995) can help students think critically and develop their cognitive and epistemological skills.

We are drawing on the theoretical work of such projects and our earlier work with representational environments in chemistry (Kozma & Russell, 1997) to develop ChemSense. Key properties of ChemSense include: (a) integration with real experiments (e.g., via probeware, spreadsheets), (b) a workspace for creating, organizing, and manipulating representations (e.g., with modeling tools, libraries of representations, authoring tools), (c) tools to scaffold critical reasoning and inquiry (e.g., annotation and linking tools, prompts to encourage reflection), and (d) mechanisms for contributing to and drawing from a knowledge community (e.g, support for uploading and downloading activities to and from a community website to publish and share work).

ChemSense is being developed on a new, innovative platform being developed by the ESCOT project (www.escot.org; Roschelle, Digiano, Pea, Kaput, 1999), using Java and related frameworks. ESCOT supports reusable, interoperable components (like graphs, tables, equation editors, simulation engines, web browsers, spreadsheets, scripting languages, many others) and easy authoring of activities that combine components to compose new lessons. Interoperable components can exchange data to create various dynamically linked representations where the same information can be simultaneously viewed and manipulated in different representational forms. Components are assembled in "activity documents" that specify the "wiring" between components (how they share data), their layout, initial values, and associated instructional and help text. Activities are saved in XML, an extensible, widely accepted, platform independent standard for saving data that is easily sharable and both human readable and machine parsable. We are co-developing and sharing components (e.g., molecule viewers, probeware interfaces), research, and evaluation with ESCOT and other groups (e.g., ChemViz), as well as building extensions to the ESCOT platform (e.g., to support whiteboard-like annotation of activity pages). We expect to complete our first set of prototype tools and activities in early fall, 1999.

Research studies

As we develop our learning environment, we are extending our research with a series of "design experiments" conducted in collaboration with teachers and students in a local (California) high school and the University of Michigan. We will both test initial versions of the environment, and use the environment to conduct experimental studies about the relationships between representational competence, epistemological beliefs and chemical understanding. The following studies are underway or planned:

• Baseline studies (underway) of teachers' and students' uses of representations and epistemological actions during discourse in high school and college chemistry courses. We are currently observing and analyzing the discourse of several students and teachers as they reason about chemical phenomena, and interviewing students about their understanding of various representations using instruments adapted from earlier work (e.g., Kozma & Russell, 1997).

• Participatory design sessions (year 2) with students and teachers using paper mockups and early working prototypes to refine the design of ChemSense and the sample activities.

• Controlled studies (years 2 and 3) to compare (a) the effects of experimenter-provided & student-self-generated representations on chemical understanding and representational competence, (b) the effects of experimenter-provided and student-self-generated epistemological prompts on students' epistemological beliefs and actions, and (c) the interaction of representations and epistemological prompts on chemical understanding.

• A year-long study (year 3) of the effect of ChemSense on representational competence, epistemological beliefs, and chemical understanding.

Summary and future directions

Our current 3-year project focus is to develop, in collaboration with teachers and students, a set of representational and annotational tools and sample activities to support the learning of chemistry in high school and college. We are also creating new partnerships to co-develop and share components, research, and evaluation. With future funding we hope to develop more "knowledge community" mechanisms——extended support for conversation between students, teachers, and mentors around student "portfolios" (including mechanisms for organizing, sharing, and commenting on student work), and multiuser facilities to support small-group collaboration both locally and at a distance (e.g., synchronous and asynchronous communication mechanisms, shared gesturing, attribution mechanisms). We view ChemSense as potential general platform and portal for chemistry education.


We thank our collaborating teachers and researchers Judy Larson, Sally Seebode, Anders Rosenquist, Christine Korbak, and Fumiko Allen for their helpful insights, comments, assistance, and support. This work is supported by a National Science Foundation Grant REC-9814653.


Gabel, D., & Bunce, D. (1994). Research on problem solving: Chemistry. In D. Gabel (Ed.), Handbook of research on science teaching and learning. New York: Macmillan.

Jackson, S., L., Stratford, S. J., Krajcik, J., & Soloway (in press). Making dynamic modeling accessible to pre-college science students. Interactive Learning Environments.

Krajcik, J. (1991). Developing students' understanding of chemical concepts. In S. Glynn, R. Yeany, & B. Britton (Eds.), The psychology of learning science (pp. 117-147). Hillsdale, NJ: Erlbaum.

Krajcik, J, Blumenfeld, P., Marx, R., Bass, K., and Fredricks, J. (1998). Middle school students initial attempts at inquiry in project-based science classrooms. Journal of the Learning Sciences, 7(3 & 4).

Kozma, R., & Russell, J. (1997). Multimedia and understanding: Expert and novice responses to different representations of chemical phenomena. Journal of Research in Science and Teaching, 43(9), 949-968.

Latour, B. (1987). Science in action. Milton Keynes, UK: Open University Press.

Moran, J., Pearson, P., Bienvenue, L., Chang, C., Nelson, D., & Pasero, S., (1997). Visualizations in teaching chemistry (Final Report). National Center for Supercomputing Applications (NCSA), University of Illinois at Urbana-Champaign.

Roschelle, J., Digiano, C., Pea, R., Kaput, J. (1998, November). Educational Software Components of Tomorrow (ESCOT). Paper presented at MSET-99, Boston, MA.

Roschelle, J., Kaput, J., & Stroup, W., (in press). SimCalc: Accelerating students' engagement with the mathematics of change. In Michael J. Jacobson & Robert B. Kozma (Eds.), Learning the sciences of the 21st century: Research, design, and implementing advanced technology learning environments. Hillsdale, NJ: Erlbaum.

Scardamalia, M., Bereiter, C., & Lamon, M. (1994). The CSILE project: Trying to bring the classroom into world 3. In Kate McGilly (Ed.), Classroom lessons: Integrating cognitive theory and classroom practice (pp. 201-228). Cambridge, MA: Bradford Books/MIT Press.

Schank, P. (1995). Computational tools for modeling and aiding reasoning: Assessing and applying the Theory of Explanatory Coherence. (Doctoral dissertation, University of California, Berkeley). (University Microfilms No. 9621352).

Slotta, J., & Linn, M. (in press). How do students make sense of Internet resources in the science classroom? In M. Jacobson & R. Kozma (Eds.), Learning the sciences of the 21st century: Research, design, and implementing advanced technology learning environments. Hillsdale, NJ: Erlbaum.

Suthers, D., Toth, E., & Weiner, A. (1997). An integrated approach to implementing collaborative inquiry in the classroom. In R. Hall, N. Miyake, & N. Enyedy (Eds.). Proceedings of the Second International Conference on Computer Support for Collaborative Learning, pp. 272-279. Toronto, Canada: University of Toronto Press.

Authors' addresses

Patricia Schank (patricia.schank@sri.com)
SRI International, Center for Technology in Learning; 333 Ravenswood Ave.; Menlo Park, CA 94025. Tel. (650) 859-3934. Fax (650) 859-3673.
Elaine Coleman (elaine.coleman@sri.com)
SRI International, Center for Technology in Learning; 333 Ravenswood Ave.; Menlo Park, CA 94025. Tel. (650) 859-6031. Fax (650) 859-4605.
Robert Kozma (robert.kozma@sri.com)
SRI International, Center for Technology in Learning; 333 Ravenswood Ave.; Menlo Park, CA 94025. Tel. (650) 859-3997. Fax (650) 859-3673.
Brian Coppola (bcoppola@umich.edu)
Department of Chemistry; University of Michigan; 930 North University; Ann Arbor, MI 48109-1055. Tel. (734) 764-7329. Fax (734) 647-4865.