A computational model for the evolution of learning physical micro-contents in peer instruction methodology

TL;DR

This paper develops a neurocognitive computational model to better understand peer instruction in physics, addressing previous phenomenological models by incorporating cognitive processes, instructor roles, and content-student interactions.

Contribution

The study introduces a new neurocognitive-based computational model that improves understanding of peer instruction dynamics in physics education.

Findings

Model successfully simulates student learning processes.

Validation with field data supports model accuracy.

Provides insights into long-term learning effects.

Abstract

One of the most important active methodologies for physics learning developed in recent years is peer instruction. Its technique has allowed, among other things, to monitor students' conceptual learning. In this sense, \textcite{PhysRevSTPER.6.020105} has proposed a model that seeks to understand the dynamics of this methodology. However, her model is very phenomenological and overlooks fundamental aspects such as the cognitive process that students follow during interaction with their peers, the role of the instructor, the connection between content and student characteristics, and long-term learning. The objective of this thesis was to develop a computational model based on neurocognitive principles that aimed to address the shortcomings found in Nitta's work. The model was formulated, simulated, and validated based on several field studies on the learning of one-dimensional graphical interpretation and the falling of bodies among science and engineering students in Bogot\'a, Colombia