A Cognitive Modeling Tutor Supporting Student Inquiry for Balancing Chemical Equations

Abstract

Improved interactive tutoring capabilities in educational software for chemistry problem solving is an important need that has been clearly articulated by teachers and students. The purpose of this work is to examine the incorporation of new concepts from the field of artificial intelligence (AI) as a route to meaningful individualized tutoring. The basic shift is to replace specific foreknowledge of problems with a direct representation of chemical and pedagogical principles and then simulate reasoning using these principles to tutor students. To assess the potential of an AI-based approach, we have developed a prototype tutorial program for balancing chemical equations that contains two important advances. First, the system can create a worked-out solution with detailed explanations for any equation entered by the student or teacher. Unlike a conventional tutorial, this is done dynamically, without the equation being stored ahead of time. Second, the program can interactively answer a variety of detailed questions about its work at each step. Studying worked-out examples plays an important role in learning, and this approach to supporting interactive student inquiry is being investigated as a method of cognitive modeling and apprenticeship intended to foster the student’s own self-explanation and question-asking abilities.