Tolerance-Based Fairness Auditing: Violation Certification and Sensitivity Screening
Abstract
As artificial intelligence is increasingly deployed, algorithmic unfairness has raised growing concerns and intensified demands for transparent fairness auditing. In practice, the tolerable degree of algorithmic unfairness depends on the specific legal, ethical, or application context. Given a prespecified tolerance threshold, an important statistical question is how to determine whether a group disparity exceeds the allowable tolerance across different auditing objectives. To address this problem, we develop a unified tolerance-based fairness auditing framework for two complementary auditing objectives: violation certification, which prioritizes control of false violation declarations, and sensitivity screening, which prioritizes reducing missed violations. For the first objective, we develop a constrained empirical likelihood test for formal settings that uses least-favorable-point calibration and can be combined with false flagging rate control for simultaneous subgroup auditing. For the second objective, we develop split empirical likelihood and adjusted split empirical likelihood tests using an adaptive boundary-proxy principle for early-warning settings. Numerical experiments show the distinct error-control--sensitivity trade-offs of these procedures. A COMPAS analysis illustrates the framework in predictive fairness auditing.
Related
- Graphene-Assisted Electrostatic Screening of Resonant Transport in MR-DWELL Dosimeters with Tunable Sensitivity--Radiation Tolerance Trade-off: A Reduced-Order Benchmark and Self-Consistent Perspective
- Conditional Mean Independence, Global Sensitivity Analysis, and Variable Screening Using nearest neighbor Graphs
- The Role of Uncertainty in Assessing the Fairness of Machine Learning Models
- Equality, Equity, and Causality in Fairness Research: A Commentary on Cheng (2026)
- Hidden Disparities in Ranking: Category-Aware Fairness via Distributional Stability
BibTeX
@misc{tang2026tolerancebased,
title = {{Tolerance-Based Fairness Auditing: Violation Certification and Sensitivity Screening}},
author = {Jie Tang and Chuanlong Xie and Lixing Zhu},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.01005},
eprint = {2610.01005},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.01005},
}
RIS
TY - UNPB
TI - Tolerance-Based Fairness Auditing: Violation Certification and Sensitivity Screening
AU - Jie Tang
AU - Chuanlong Xie
AU - Lixing Zhu
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.01005
UR - https://arxiv.org/abs/2610.01005
AB - As artificial intelligence is increasingly deployed, algorithmic unfairness has raised growing concerns and intensified demands for transparent fairness auditing. In practice, the tolerable degree of algorithmic unfairness depends on the specific legal, ethical, or application context. Given a prespecified tolerance threshold, an important statistical question is how to determine whether a group disparity exceeds the allowable tolerance across different auditing objectives. To address this problem, we develop a unified tolerance-based fairness auditing framework for two complementary auditing objectives: violation certification, which prioritizes control of false violation declarations, and sensitivity screening, which prioritizes reducing missed violations. For the first objective, we develop a constrained empirical likelihood test for formal settings that uses least-favorable-point calibration and can be combined with false flagging rate control for simultaneous subgroup auditing. For the second objective, we develop split empirical likelihood and adjusted split empirical likelihood tests using an adaptive boundary-proxy principle for early-warning settings. Numerical experiments show the distinct error-control--sensitivity trade-offs of these procedures. A COMPAS analysis illustrates the framework in predictive fairness auditing.
ER -