Kenneth Forbus, Institute for the Learning Sciences, Northwestern University
Marcia Linn, University of California at Berkeley
The opportunity: We would like to explore the role of self-explanatory simulators in teaching middle-school science. Self-explanatory simulators combine qualitative, conceptual explanations with numerical simulation. For example, a self-explanatory simulator can produce plots of the numerical values a parameter takes, like any traditional numerical simulator. But it can also summarize the behavior of the system in terms of conceptual events (i.e., liquid in a pot heating up for a while and then starting to boil) and explain the causal relationships that hold between the parameters (i.e., an increase in heat causes a change in temperature, until the boiling point is reached). We believe that this technology could be extremely valuable for education. We propose to explore this possibility by creating a set of simulation activities to complement an existing curriculum.
The curriculum we are planning to use is from Linn and her colleagues at Berkeley, who have developed a middle-school curriculum for learning about heat and temperature. Their methods are especially interesting because results from longitudinal studies suggest that what is learned is retained better than standard teaching methods. We plan on creating a complementary set of simulation-based laboratory activities, such as boiling fluids in various containers, so students can explore aspects of the phenomena that would otherwise be too dangerous, expensive, or tedious to do in real-time with physical systems. (Example: A student once suggested on a usenet group that they were about to create a boiler for a model steam engine out of a soda can. Dozens of subsequent postings explained, with great urgency, why they should not do this experiment in real life.)
What we will do: We will create a suite of activities using self-explanatory simulators for helping students learn about thermal phenomena. These activities, along with a modified version of Linn's curriculum, will become the starting point for a Work Circle (involving NWU and Michigan researchers with personnel from the Chicago Public school system, as part of the LIS center here) for the development of a curricular unit which would be used in the Chicago Public School system, starting in the 1999-2000 school year. We will also make these activities and materials available through the center, for more widespread experimentation and use.
What we need: The technology development of self-explanatory simulators is already under separate funding, from NASA Ames. This includes developing domain theories and an integrated development environment that could be directly used in this project. We are asking for summer support for one Learning Sciences graduate student to develop these activities.