Opinion dynamics model of collaborative learning

TL;DR

This paper introduces a simple opinion dynamics model to understand how group discussions lead to correct answers, emphasizing the roles of diversity, memory, and group size in collaborative learning.

Contribution

The paper presents a novel opinion dynamics model incorporating influence, diversity, and inertia, providing insights into effective group sizes and initial opinion distributions for collaborative learning.

Findings

Diverse initial opinions improve group performance.

Lower memory capacity accelerates consensus formation.

Small groups of three to four members perform better.

Abstract

We propose a simple model to explore an educational phenomenon where the correct answer emerges from group discussion. We construct our model based on several plausible assumptions: (i) We tend to follow peers' opinions. However, if a peer's opinion is too different from yours, you are not much influenced. In other words, your opinion tends to align with peers' opinions, weighted by the similarity to yours. (ii) Discussion among group members helps the opinion to shift toward the correct answer even when the group members do not know it clearly. However, if everyone tells exactly the same, you often get lost and it becomes more difficult to find the correct answer. In other words, you can find the correct answer when everyone has largely different voices. (iii) We are sometimes stuck to our past. If you keep one opinion for a long time, such a memory works like an inertia in classical mechanics. We use our model to perform numerical investigations and find that the performance of a group is enhanced when initial opinions are diverse, that a lower memory capacity makes consensus occur faster, and that a small group size, typically three or four, is beneficial for better group performance.