| Project | ChemSense |
| Contact | Robert Kozma |
| rkozma@unix.sri.com | |
| URL | http://www.sri.com/policy/ctl/html/chemsense.html |
| Project description | We currently have a proposal under consideration which, if funded, would allow us to develop a computational environment to support students’ social construction of understanding in chemistry. Entitled ChemSense, the environment would allow students to use real data from experiments they conduct in the chemistry laboratory and enter these data (either manually or through probeware) into a computational model of the phenomenon. This model would, in turn, drive multiple representations of the phenomenon (such as dynamic graphs, animations, etc.). Students could then manipulate the model and annotate these representations, using them to construct explanations for the phenomena they explored in the laboratory. |
| Theoretical background | Our current work is informed earlier work on the use of representations and explanations to support understanding in science. Our work has a social constructivist theoretical orientation in which the discourse of collaborators is supported by the representational resources and instructional scaffolding embedded in the learning environment.In our early projects, we examined the relationship between representation and understanding with chemistry experts (i.e., chemistry graduate students and faculty) and novices (i.e., college chemistry students). Our findings, recently published in the Journal of Research in Science Teaching (Kozma & Russell, 1997), show that experts are much better than novices in using a range of representations (chemical equations, graphs, video segments of experiments, and molecular animations) to create large, chemically meaningful clusters. They describe these clusters using conceptual terms, such as “gas law,” and “collision theory.” Novices create small clusters using single representational forms and their descriptions are based on the surface features of the representations, such as color, objects depicted, graph labels, and types of representations (e.g., “red molecules bouncing around,” “graphs of pressure and concentration,” etc.).In related studies (Kozma, Russell, Jones, Marx, & Davis, 1996; Kozma, in press), we explored how students used simulation software that employs multiple linked representations to understand chemical phenomena. We show how the literal features of these representations support the discourse of chemistry students working in pairs as they collaborate to construct an understanding of chemical phenomena they are simulating.In other studies, we have examined student reasoning and discussion as it is guided with "scaffolds" in the form of computer prompts that they use while working collaboratively to solve science problems. These prompts provide students with suggestions or procedures that facilitate their ability to reason more "scientifically" by provoking explanations and justifications for their responses to problems (Coleman, 1995; 1996). |
| Challenges | In our research, we are concerned with the specific features of representations and kinds of scaffolding that make perceptually inaccessible scientific phenomena (e.g., molecular bonding and reactions) conceptually accessible to students. We are also interested in studying how students use models/visualizations to reason through scientific problems in chemistry and whether this kind of activity enables students to acquire a more sophisticated sense of models (i.e., as representational entities that can both represent and test underlying ideas). Our proposed research would address these questions.Technologically, we are concerned with creating tools and computational models that would allow us to generate a range of dynamic representations that would be driven by student data. Particularly problematic are molecular-level animations that show the composition of compounds and the breaking and forming of bonds with some degree of fidelity, vis-à-vis the actual chemistry. As difficult as it is to create such representations on the fly, they are likely to be crucial in helping students understand the underlying chemical phenomena. |
| Partnership | We would like to continue our collaboration with chemists who are interested in exploring students’ understanding of chemistry and designing visual models that support students’ collaborative learning. We are also interested in technological partners who can work with us to design advanced computational environments that generate a range of real time, dynamic representations and address the technological concerns mentioned above. |