Further Results from the Evaluation of an Intelligent Computer Tutor to Coach Self-Explanation

We present further results on the educational effectiveness of an intelligent computer tutor that helps students learn effectively from examples by coaching self-explanation – the process of explaining to oneself an example worked-out solution. An earlier analysis of the results from a formative evaluation of the system provided suggestive evidence that it could improve students’ learning. In this paper, we present additional results derived from a more comprehensive analysis of the experimental data. They provide a stronger indication of the system’s effectiveness and suggest general guidelines for effective support of self-explanation during example studying.