Algorithmic fairness in education

Алгоритмическая справедливость в образовании
René F. Kizilcec, Hansol Lee, Hansol Lee
2022-08-11

algorithmic biasalgorithmic discriminationalgorithmic fairnesscausal fairnesseducational predictive models
Data-driven predictive models are increasingly used in education to support students, instructors, and administrators, which has raised concerns about the fairness of their predictions and uses of these algorithmic systems. In this introduction to algorithmic fairness in education, we draw parallels to prior literature on educational access, bias, and discrimination, and we examine core components of algorithmic systems (measurement, model learning, and action) to identify sources of bias and discrimination in the process of developing and deploying these systems. Statistical, similarity-based, and causal notions of fairness are reviewed and contrasted in how they apply in educational contexts. Recommendations for policymakers and developers of educational technology offer guidance for promoting algorithmic fairness in education.
1
Algorithmic predictive systems in education raise fairness concerns because their predictions and downstream uses can reproduce bias and discrimination.
2
Bias may enter educational algorithmic systems through measurement, model learning, and subsequent actions based on predictions.
3
Recommendations for policymakers and educational technology developers provide guidance for promoting fairness throughout system development and deployment.
4
Statistical, similarity-based, and causal fairness definitions differ in their applicability and implications for educational contexts.
5
The paper connects algorithmic fairness in education with established concerns about educational access, bias, and discrimination.

Data-driven predictive models and algorithmic systems used in education

Sources and notions of bias, discrimination, and fairness in the measurement, learning, and deployment of educational algorithmic systems

Publication Details
Publication Date
2022-08-11
Journal
Publisher
ISSN
Cited by
344
Access Type
Author Information
Authors
René F. Kizilcec
Hansol Lee
Hansol Lee
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%