2025 ASEE Annual Conference & Exposition

IUSE:HSI: A NetLogo-based Approach to Predictive Estimation of Student Performance and Corresponding Tutoring Demand within STEM Undergraduate Courses

Presented at NSF Grantees Poster Session I

Tutoring support is vital to eliminate knowledge gaps and achieve learning outcomes, while attaining instructional scalability to large class sizes common in STEM courses. However, determining which courses require additional tutoring support is challenging due to the lack of formal quantitative measurement tools, thus hindering the ideal provision of Teaching Assistant (TA) allocation. Herein, we develop a NetLogo framework for an Agent-Based Model (ABM) designed to simulate student progress and performance in a required Electrical and Computer Engineering (ECE) undergraduate course. It predicts and quantifies course outcomes under varying amounts of tutoring support via TA office hours. The ABM incorporates key parameters such as learners' attainment in previous semesters, grading schemata, and tutoring impact to predict the corresponding number of at-risk students. Meanwhile, the technical approach utilizes a NetLogo-based implementation of the ABM, which realizes a flexible, modular, and observable design.

The ABM is composed of three primary components: student agents which represent individual students along with their content proficiency, a course environment which encapsulates the grading scheme along with tutoring support parameters of the TAs available, and a feedback mechanism which enables the ABM to adjust its predictions based on the instructor's input. The technical results demonstrate that the ABM is capable of accurately predicting the number of students having significant risk of DFW, i.e. earning a course letter grade of D or F, or withdrawing (W). The results show that the ABM is robust and reliable in its predictions of DFW rate, whereas three out of four semester configurations analyzed indicated that the ABM’s predicted values bordered the 95% confidence interval. When measuring accuracy, test runs included course enrollment ranging from 70 to 121 students, commensurate with actual course delivery enrollments. The NetLogo model was parameterized towards attaining the worthy objective of lowering the DFW rate. Furthermore, to assist administrative decision making, it computes the monetary cost of tutoring per supported student. This new metric, known as Remediation Cost Per Supported Student (RCSS), delivers a quantitative measurement of cost-effectiveness for a course staffing configuration when considering the number of tutors paid and the number of students who received remediation. The model's performance is evaluated through a series of experimental scenarios, which involve varying student enrollment, grading schemes, and teaching assistant support levels using a dataset of previous course offerings.

Authors

Mr. Paul Amoruso University of Central Florida [biography]

Paul Amoruso holds a Bachelor of Science and Master of Science degree in Computer Engineering from the Department of Electrical Engineering and Computer Science at the University of Central Florida. He is currently pursuing a Doctor of Philosophy degree in Computer Engineering at the same institution. Since 2022, he has held the positions of Graduate Research Assistant and Graduate Teaching Assistant within the Department of Electrical Engineering and Computer Science at the University of Central Florida. Amoruso’s research interests encompass advancements in machine learning implementations, particularly in the domains of education, technology-enabled learning, and feedback-driven grading approaches.
Ivan Garibay University of Central Florida
Dr. Joel Alejandro Mejia http://orcid.org/0000-0003-3908-9930 University of Cincinnati [biography]

Dr. Joel Alejandro (Alex) Mejia is a Professor of Engineering Education in the Department of Engineering &Computing Education at the University of Cincinnati. Prior to working in academia, he worked as a project engineer for the Department of Defense, Rio Tinto, and FLSmidth Minerals. Dr. Mejia’s work seeks to analyze and describe the assets, tensions, contradictions, and cultural collisions many Latino/a/x students experience in engineering through pláticas and testimonios. His research has contributed to the integration of critical theoretical frameworks in engineering education to investigate deficit ideologies and their impact on minoritized communities. He is particularly interested in approaches that contribute to a more expansive understanding of engineering in sociocultural contexts, engineering education in bilingual contexts, and the impact of critical consciousness in engineering practice.
Dr. Laurie O Campbell http://orcid.org/0000-0001-7313-5457 University of Central Florida [biography]

Laurie O. Campbell, Ed.D., is an Assoc. Professor of STEM and Instructional Design and Technology, at the University of Central Florida. She pursues research related to STEM curriculum and STEM identity among underserved and underrepresented populations, p
florencio Eloy Hernandez Texas A&M University - Corpus Christi
Dr. Ronald F. DeMara P.E. http://orcid.org/0000-0001-6864-7255 University of Central Florida [biography]

Ronald F. DeMara is Pegasus Professor in the Department of Electrical and Computer Engineering, and joint faculty member of Computer Science, at the University of Central Florida, where he has been a full-time faculty member since 1993. He has completed over 325 articles, 50 funded projects as PI or Co-PI, and 56 graduates as Ph.D. dissertation and/or M.S. thesis advisor. He was previously an Associate Engineer at IBM and a Visiting Research Scientist at NASA Ames, in total for four years, and has been a registered Professional Engineer since 1992. He has served ten terms as a Topical Editor or Associate Editor of various IEEE Transactions and in many IEEE/ACM/ASEE conferences including General Co-Chair of GLSVLSI-2023. He has received the Joseph M. Biedenbach Outstanding Engineering Educator Award from IEEE and is a Fellow of AAAS.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 22, 2025, and to all visitors after the conference ends on June 25, 2025

« View session

For those interested in:

computer science
engineering
Faculty
information technology
undergraduate