TY - GEN

T1 - On (assessing) the fairness of risk score models

AU - Petersen, Eike

AU - Ganz, Melanie

AU - Holm, Sune

AU - Feragen, Aasa

N1 - Publisher Copyright: © 2023 ACM.

PY - 2023

Y1 - 2023

N2 - Recent work on algorithmic fairness has largely focused on the fairness of discrete decisions, or classifications. While such decisions are often based on risk score models, the fairness of the risk models themselves has received considerably less attention. Risk models are of interest for a number of reasons, including the fact that they communicate uncertainty about the potential outcomes to users, thus representing a way to enable meaningful human oversight. Here, we address fairness desiderata for risk score models. We identify the provision of similar epistemic value to different groups as a key desideratum for risk score fairness, and we show how even fair risk scores can lead to unfair risk-based rankings. Further, we address how to assess the fairness of risk score models quantitatively, including a discussion of metric choices and meaningful statistical comparisons between groups. In this context, we also introduce a novel calibration error metric that is less sample size-biased than previously proposed metrics, enabling meaningful comparisons between groups of different sizes. We illustrate our methodology - which is widely applicable in many other settings - in two case studies, one in recidivism risk prediction, and one in risk of major depressive disorder (MDD) prediction.

AB - Recent work on algorithmic fairness has largely focused on the fairness of discrete decisions, or classifications. While such decisions are often based on risk score models, the fairness of the risk models themselves has received considerably less attention. Risk models are of interest for a number of reasons, including the fact that they communicate uncertainty about the potential outcomes to users, thus representing a way to enable meaningful human oversight. Here, we address fairness desiderata for risk score models. We identify the provision of similar epistemic value to different groups as a key desideratum for risk score fairness, and we show how even fair risk scores can lead to unfair risk-based rankings. Further, we address how to assess the fairness of risk score models quantitatively, including a discussion of metric choices and meaningful statistical comparisons between groups. In this context, we also introduce a novel calibration error metric that is less sample size-biased than previously proposed metrics, enabling meaningful comparisons between groups of different sizes. We illustrate our methodology - which is widely applicable in many other settings - in two case studies, one in recidivism risk prediction, and one in risk of major depressive disorder (MDD) prediction.

KW - Algorithmic fairness

KW - Calibration

KW - Ethics

KW - Major depressive disorder

KW - Ranking

KW - Recidivism

KW - Risk scores

U2 - 10.1145/3593013.3594045

DO - 10.1145/3593013.3594045

M3 - Article in proceedings

AN - SCOPUS:85163674977

T3 - ACM International Conference Proceeding Series

SP - 817

EP - 829

BT - Proceedings of the 6th ACM Conference on Fairness, Accountability, and Transparency, FAccT 2023

PB - Association for Computing Machinery, Inc.

T2 - 6th ACM Conference on Fairness, Accountability, and Transparency, FAccT 2023

Y2 - 12 June 2023 through 15 June 2023

ER -