Knowledge construction with causal concept maps in a teachable agent environment

Abstract

We have developed Betty's Brain [1], a computer-based learning environment that employs the learning-by-teaching paradigm to foster students' acquisition of science knowledge and self-regulated learning strategies. The system provides students with opportunities for self-directed, open-ended learning in science. In this learning environment, students are given a knowledge construction task in which they teach a virtual agent by engaging in an iterative process of reading source material and structuring their knowledge in a causal concept map for a particular science domain (e.g., ecology or thermo-regulation). The agent, then, can use this map to answer questions and take quizzes. © 2011 Springer-Verlag Berlin Heidelberg.