Using fair AI to predict students’ math learning outcomes in an online platform

ABSTRACT

As instruction shifts away from traditional approaches, online learning has grown in popularity in K-12 and higher education. Artificial intelligence (AI) and learning analytics methods such as machine learning have been used by educational scholars to support online learners on a large scale. However, the fairness of AI prediction in educational contexts has received insufficient attention, which can increase educational inequality. This study aims to fill this gap by proposing a fair logistic regression (Fair-LR) algorithm. Specifically, we developed Fair-LR and compared it with fairness-unaware AI models (Logistic Regression, Support Vector Machine, and Random Forest). We evaluated fairness with equalized odds that caters to statistical type I and II errors in predictions across demographic subgroups. The results showed that the Fair-LR could generate desirable predictive accuracy while achieving better fairness. The findings implied that the educational community could adopt a methodological shift to achieve accurate and fair AI to support learning and reduce bias.

KEYWORDS:

• fair AI
• learning analytics
• online learning
• machine learning

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The research reported here was supported by the Institute of Education Sciences, US Department of Education, through Grant R305C160004 to the University of Florida, the University of Florida AI Catalyst Grant -P0195022, and the University of Florida Informatics Institute Seed Funding. The opinions expressed are those of the authors and do not represent the views of the University of Florida, Institute of Education Sciences, or those of the US Department of Education.