Paper Authors

Abstract

NOTE: The first page of text has been automatically extracted and included below in lieu of an abstract

A Case Study of Multi-Agent-Based Simulation in Undergraduate Materials Science Education

1. Introduction

This paper reports on a user study of a computer-aided learning environment for college- level Materials Science. “MaterialSim” is an agent-based set of microworlds built by the authors within the NetLogo1 modeling environment, for investigating phenomena such as crystallization, solidification, grain growth and annealing. This design started out from a literature review on Materials Science education, extensive classroom observations and analysis of class materials. This preliminary phase suggested that students’ understanding of the subject matter was problematic, and that the teaching strategies and technological learning resources were not up to the challenge of the very complex content being taught. Based on this preliminary diagnosis, we created a set of exploratory computer microworlds and designed a user study for evaluating its effectiveness. A total of seventeen undergraduate students enrolled in an introductory Materials Science course participated in the study in 2004 and 2005, which was comprised of a survey, pre-interview, interaction with the computer models, and model building.

The complexity of college-level content in the field of Materials Science is such that teachers oftentimes end up resorting to a multitude of aggregate-level models to explain a particular phenomenon, each of them requiring a different set of equations. Our first research question was to investigate the consequences of this common teaching strategy: what kind of understanding did this multitude of explanation models foster in students?

MaterialSim’s multi-agent-based computer models were built within a different framework. Instead of resorting to a variety of aggregate-level description, the core feature of the multi-agent-based perspective is the application of a small number of local rules to capture fundamental causality structures underlying complex behaviors within a domain. The rationale for the design is that this agent-based perspective may foster different, and perhaps more generative understanding of the relevant scientific phenomena. Instead of multiple models or numerous equations, this framework focuses on a small number of elementary models which can be applied to a variety of scientific phenomena. This is in opposition to common teaching strategies in Materials Science. Our second research question, thus, was to study students’ understanding of the Materials Science content departing from this agent-based perspective.

A third research question, about which we will present only preliminary results, was to investigate the learning outcomes of building computer models. In other words, does coding (i.e., programming) multi-agent-based models generate deeper understanding of the scientific phenomena, as opposed to just interacting with ready-made models?

To answers our research questions, we present evidence in the form of excerpts and samples of students’ work, which demonstrates that the experience with MaterialSim enabled them to identify and understand some of the unifying principles in Materials Science and build new models based on those principles.

• Citation

• Format
  • APA
  • APA - LaTeX bibitem
  • MLA
  • MLA - LaTeX bibitem
  • Bibtex
  • EndNote - RIS

Blikstein, P., & Wilensky, U. (2006, June), A Case Study Of Multi Agent Based Simulation In Undergraduate Materials Science Education Paper presented at 2006 Annual Conference & Exposition, Chicago, Illinois. 10.18260/1-2--1396