Grant Open

Fair Prediction of College-Student Success Using Multivariate Adaptive Regression Splines

NCER
Program: Statistical and Research Methodology in Education
Program topic(s): Early Career
Award amount: $299,461
Principal investigator: Hadis Anahideh
Awardee:
University of Illinois, Chicago
Year: 2022
Project type:
Methodological Innovation
Award number: R305D220055

Purpose

The purpose of this project is to develop a new fair Multivariate Adaptive Regression Splines (MARS) statistical model to predict college-student success. MARS is a parsimonious non-parametric regression model that can identify useful input variables through a built-in feature-selection step when many potential variables are considered. MARS also renders an easily interpretable model, making it more helpful for use in higher education settings.

Project Activities

The research team will develop the model for this project by decorrelating sensitive and non-sensitive attributes and incorporating fairness as a constraint in the optimization step of the MARS algorithm. They will use a proactive auditing approach to assess and calibrate fairness of the model. After developing the model, the researchers will use it on data from the Education Longitudinal Study of 2002 (ELS:2002) and Integrated Postsecondary Education Data System (IPEDS) to evaluate its effectiveness at predicting students' likelihood of success.

People and institutions involved

IES program contact(s)

Charles Laurin

Education Research Analyst
NCER

Products and publications

Products: The researchers will also develop and post to Github open-source, user-friendly software in Python with a comprehensive user guide for researchers and education practitioners. The team will also disseminate their work at conferences and through peer-reviewed journal manuscripts.

ERIC Citations: Find available citations in ERIC for this award here.

Publications:

Anahideh, H., Nezami, N., & Asudeh, A. (2021). Finding Representative Group Fairness Metrics Using Correlation Estimations. arXiv preprint arXiv:2109.05697.

Di Carlo, F., Nezami, N., Anahideh, H., & Asudeh, A. (2023). FairPilot: An Explorative System for Hyperparameter Tuning through the Lens of Fairness. arXiv preprint arXiv:2304.04679.

Gándara, D., Anahideh, H., Ison, M. P., & Tayal, A. (2023). Inside the Black Box: Detecting and Mitigating Algorithmic Bias across Racialized Groups in College Student-Success Prediction. arXiv preprint arXiv:2301.03784.

Haghighat, P., Gándara, D., Kang, L., & Anahideh, H. (2024, March). Fair Multivariate Adaptive Regression Splines for Ensuring Equity and Transparency. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 38, No. 20, pp. 22076-22086).

Nezami, N., Haghighat, P., Gándara, D., & Anahideh, H. (2024). Assessing Disparities in Predictive Modeling Outcomes for College Student Success: The Impact of Imputation Techniques on Model Performance and Fairness. Education Sciences, 14(2), 136.

Supplemental information

Co-Principal Investigator: Gándara, Denisa

Questions about this project?

To answer additional questions about this project or provide feedback, please contact the program officer.

 

Questions about this project?

To answer additional questions about this project or provide feedback, please contact the program officer.

 

You may also like