Exploring the Viability of Agent-Based Modeling to Extend Qualitative Research: Comparison of Computational Platforms

Samantha Splendido; Catherine G. P. Berdanier

Download Paper | Permalink

Conference: 2023 ASEE Annual Conference & Exposition
Location: Baltimore , Maryland
Publication Date: June 25, 2023
Start Date: June 25, 2023
End Date: June 28, 2023
Conference Session: Investigating Student Pathways to and through Undergraduate and Graduate Programs
Tagged Division: Educational Research and Methods Division (ERM)
Page Count: 17
DOI: 10.18260/1-2--43658
Permanent URL: https://peer.asee.org/43658
Download Count: 188

Request a correction

Paper Authors

biography

Samantha Splendido Pennsylvania State University, University Park

visit author page

Sam Splendido is a Ph.D. student in Mechanical Engineering at Pennsylvania State University. She is currently a graduate research assistant under Dr. Catherine Berdanier in the Engineering Cognitive Research Laboratory (ECRL). She earned her B.S. in Biomedical and Mechanical Engineering from Pennsylvania State University.

visit author page

biography

Catherine G. P. Berdanier Pennsylvania State University orcid.org/0000-0003-3271-4836

visit author page

Catherine G.P. Berdanier is an Assistant Professor in the Department of Mechanical Engineering at Pennsylvania State University. She earned her B.S. in Chemistry from The University of South Dakota, her M.S. in Aeronautical and Astronautical Engineering and her PhD in Engineering Education from Purdue University. Her research expertise lies in characterizing graduate-level attrition, persistence, and career trajectories; engineering writing and communication; and methodological development.

visit author page

Download Paper | Permalink

Abstract

The purpose of this methods paper is to identify the opportunities and applications of agent-based modeling (ABM) methods to interpretative qualitative and educational research domains. The context we explore in this paper considers graduate engineering attrition, which has been a funded research focus of our group for ten years. In attrition research, as with all human research, it is impossible and unethical to imperil real graduate students by subjecting them to acute stressors that are known to contribute to attrition in order to “test” different combinations of factors on persistence and attrition. However, ABM methods have been applied in other human decision-making contexts in which a computer applies researcher-programmed logic to digital actors, invoking them to make digital decisions that mimic human decision-making. From our research team’s ten years of research studying graduate socialization and attrition and informed from a host of theories that have been used in literature to investigate doctoral attrition, this paper compares the utility of two programming languages, Python and NetLogo, in conducting agent-based modeling to model graduate attrition as a platform. In this work we show that both platforms can be used to simulate attrition and persistence scenarios for thousands of digital agent-students simultaneously to produce results that agree with both with previous qualitative data and that agree with aggregate attrition and persistence statistics from literature. The two languages differ in their integrated development environments (IDE) with the methods of producing the models customizable to fit the needs of the study. Additionally, the size of the intended agent pool impacted the efficiency of the data collection. As computational methods can transform educational research, this work provides both a proof-of-concept and recommendations for other researchers considering employing these methods with these and similar platforms. Ultimately, while there are many programming languages that can perform agent-based modeling tasks, researchers are responsible for translating high quality, theory-driven, interpretive research into a computational model that can model human decision-making processes.

Citation
Format

Splendido, S., & Berdanier, C. G. P. (2023, June), Exploring the Viability of Agent-Based Modeling to Extend Qualitative Research: Comparison of Computational Platforms Paper presented at 2023 ASEE Annual Conference & Exposition, Baltimore , Maryland. 10.18260/1-2--43658

TY  - CPAPER
AB  - The purpose of this methods paper is to identify the opportunities and applications of agent-based modeling (ABM) methods to interpretative qualitative and educational research domains. The context we explore in this paper considers graduate engineering attrition, which has been a funded research focus of our group for ten years. In attrition research, as with all human research, it is impossible and unethical to imperil real graduate students by subjecting them to acute stressors that are known to contribute to attrition in order to “test” different combinations of factors on persistence and attrition. However, ABM methods have been applied in other human decision-making contexts in which a computer applies researcher-programmed logic to digital actors, invoking them to make digital decisions that mimic human decision-making. From our research team’s ten years of research studying graduate socialization and attrition and informed from a host of theories that have been used in literature to investigate doctoral attrition, this paper compares the utility of two programming languages, Python and NetLogo, in conducting agent-based modeling to model graduate attrition as a platform. In this work we show that both platforms can be used to simulate attrition and persistence scenarios for thousands of digital agent-students simultaneously to produce results that agree with both with previous qualitative data and that agree with aggregate attrition and persistence statistics from literature. The two languages differ in their integrated development environments (IDE) with the methods of producing the models customizable to fit the needs of the study. Additionally, the size of the intended agent pool impacted the efficiency of the data collection. As computational methods can transform educational research, this work provides both a proof-of-concept and recommendations for other researchers considering employing these methods with these and similar platforms. Ultimately, while there are many programming languages that can perform agent-based modeling tasks, researchers are responsible for translating high quality, theory-driven, interpretive research into a computational model that can model human decision-making processes.

AU  - Samantha Splendido
AU  - Catherine G. P. Berdanier
CY  - Baltimore , Maryland
DA  - 2023/06/25
PB  - ASEE Conferences
TI  - Exploring the Viability of Agent-Based Modeling to Extend Qualitative Research: Comparison of Computational Platforms
UR  - https://peer.asee.org/43658
DO  - 10.18260/1-2--43658
ER  -