Tracking the Course of Mathematical Learning

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

The goal of this project is to improve our ability to track how students learn to solve mathematical problems and how better to instruct mathematics in the context of intelligent tutoring systems. The underlying model of cognition informing the research program is Anderson's ACT-R. The researchers will look both at the potential of eye tracking and of brain imaging to improve the tracking of learning. It will focus on a computer-based algebra tutor that is currently in use in high school and is being adapted for use in middle schools. Past research has shown that instructional interventions based on eye movements can improve learning by detecting when students are having problems and by disambiguating the nature of their confusions. The current research will focus on developing interventions based on using a practical web-based camera that might be deployed in the classroom. The goal is to devise a system that will respond to the course of learning in the individual. They propose three project lines: The first project, involving fMRI brain imaging, will identity brain correlates of critical aspects of mathematics learning (e.g., whether there is a different pattern of activiation when students have simply made a computational error as opposed to when are they confused). The second project will conduct brain imaging studies during tutorial interactions. The research team will attempt to confirm that the Regions of Interest identified in the first line will generalize to learning with the tutor. They will also attempt contingent tutoring in which they branch instructional decisions depending on the brain signal. The third project will focus on advancing eye tracking methodology to the point where it can be used for instruction in the classroom by means of a web-cam.