Positive and negative verbal feedback for Intelligent Tutoring Systems

Abstract. We built three different versions of an ITS on a letter pattern extrapolation task: in one version, students only receive color-coded feedback; in the second, they receive verbal feedback messages when they perform correct actions, and in the third, when they make a mistake. We found that time on task and number of errors are predictive of performance on the post-test rather than the type of feedback. Keywords. Intelligent Tutoring Systems. Natural Language feedback.

• Type: Conference
• Year: 2005
• Where: AIED
• Authors: Barbara Di Eugenio, Xin Lu, Trina C. Kershaw, Andrew Corrigan-Halpern, Stellan Ohlsson
• Keywords: AIED 2005, Artificial Intelligence, Color-coded Feedback, Pattern Extrapolation Task, Verbal Feedback Messages