Woolf, B., Reid, J., Stillings, N., Bruno, M., Murray, D., Reese, P., Peterfreund, A. & Rath, K. “A General Platform for Inquiry Learning.” Proceedings of the International Conference on Intelligent Tutoring Systems, Biarritz, France, June 2002. A General Platform for Inquiry Learning Beverly Park Woolf, + John Reid,* Neil Stillings,* Merle Bruno,* Dan Murray # Paula Reese, + Alan Peterfreund, ++ Kenneth Rath, + + University of Massachusetts, *Hampshire College # University of Rhode Island, ++ Peterfreund Associates Abstract. We have built and evaluated software that supports students to reason about a phenomenon and develop hypotheses to explain it. The goal is for students to ask their own questions, engage in hypothesis generation and make and test predictions about theories. Intelligent tutoring and a discovery approach guide students’ inquiry in these problem-based modules. The inquiry infrastructure provides a base to make inquiry-oriented instruction more widely available. We are expanding this model across three domains, several institutions and teaching style. 1. Why is Inquiry Learning Difficult? One great challenge for intelligent tutors is to keep students engaged in authentic and active work. Processes, such as the origins of rivers, medical diagnosis, or equipment failure, have always been difficult to convey via text or traditional lecture, yet conceptual understanding of the process is key to understanding. Inquiry-oriented learning focused on processes has proven effective in dedicated small- college settings such as Hampshire College (Stillings et al., 1999, 2000; D’Avanzo & McNeal, 1997). Moving it to large post-secondary classrooms is too labor intensive. Based on prior NSF research, 1 we found that instructors devoted significantly more time to activities designed for the acquisition of inquiry skills and that these practices led to greater changes in students' inquiry skills and epistemologies. Inquiry in either a classroom or supported by computers is very difficult. Teachers in a typical classroom ask 95% of the questions, mostly requiring short answers (Graesser & Person, 1994; Hmlo-Silver, 2002). Unfortunately, most computer-assisted teaching systems also are structured toward traditional delivery of concepts, facts and findings: they direct students to a single correct answer and often push them further away from inquiry-oriented instruction. Figure 1: View of a bend in the Tuolumne River, Yosemite National Park with observation pad beneath for students to enter detailed observations 1 NSF KDI REC 9729363 Inquiry Based Science Education: Cognitive Measures and System Support.