A Multi-Agent-Based Approach for the Study of Student Behavior Dynamics in Peer Learning Environments

Publication Name

TALE 2021 - IEEE International Conference on Engineering, Technology and Education, Proceedings


Peer learning indicates an educational environment where students teach and learn from each other. Understanding how students' peer learning behaviors change over time would bring benefits such as providing clues to teachers to encourage peer learning. Studying such behavior dynamics in realistic environments could be time-consuming due to the cost of human labor. This paper proposes to use computer simulations as an efficient way to study student behavior dynamics in peer learning environments. Specifically, a student is modeled as an autonomous agent who acts as either a tutor or a learner in peer learning interactions. An interaction is modeled in a game-theoretic framework where the immediate effects of interacting agents' joint behaviors are specified. The interaction results would influence the behaviors of agents based on their behavior choice strategies. Three typical strategies are presented and tested in experiments. Experiment results reveal some insights into students' behavior dynamics under various settings, which might help teachers to find ways to encourage peer learning among students.

Open Access Status

This publication is not available as open access

First Page


Last Page


Funding Number


Funding Sponsor

Australian Research Council



Link to publisher version (DOI)