Effort-based Feedback in Intelligent Tutoring Systems
Many intelligent tutoring systems employ “peformance-based feedback” that targets students’ problem-solving performance (i.e., whether a student-given input is correct or not). In the actual classroom, however, teachers give not just performance-based feedback but also “effort-based feedback” that addresses students’ effort, regardless of their problem-solving performance (e.g., “You are doing great, keep up the good work!”).
We designed an intelligent tutoring system for elementary math where the system gives learners a variety of effort-based feedback while they solve problems. A classroom study in a Japanese elementary school found that the system with effort-based feedback enhanced student learning outcomes (from pretest to posttest). When compared against a version with traditional performance-based feedback, there was no difference in terms of learning outcomes between the two version of the system.
Project members: Shintaro Sato (lead), Prof. Dr. Melanie Platz, and Tomo Nagashima
Relevant publications
- Sato, S., Platz, M., & Nagashima, T. (2024). Which leads to more effective learning in intelligent tutoring software, effort-based or performance-based feedback? In Proceedings of the 64th Annual Meeting of the Cognitive Science Society, Rotterdam, Netherlands. Cognitive Science Society. [link]