# Fairness criteria through the lens of directed acyclic graphical models

arXiv: https://arxiv.org/abs/1906.11333 · HTML: https://mlsanity.com/paper/1906.11333/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Jun 2019.

## Built on by (most cited first, 6 of 6)

- [Explaining Algorithmic Fairness Through Fairness-Aware Causal Path Decomposition](https://mlsanity.com/paper/2108.05335/) (2021)
- [Gradual (In)Compatibility of Fairness Criteria](https://mlsanity.com/paper/2109.04399/) (2021)
- [Fair CoVariance Neural Networks](https://mlsanity.com/paper/2409.08558/) (2024)
- [A Review of Fairness and A Practical Guide to Selecting Context-Appropriate Fairness Metrics in Machine Learning](https://mlsanity.com/paper/2411.06624/) (2024)
- [Learning Interpretable Fair Representations](https://mlsanity.com/paper/2406.16698/) (2024)
- [Causal foundations of bias, disparity and fairness](https://mlsanity.com/paper/2207.13665/) (2022)

## Abstract (arXiv)

A substantial portion of the literature on fairness in algorithms proposes, analyzes, and operationalizes simple formulaic criteria for assessing fairness. Two of these criteria, Equalized Odds and Calibration by Group, have gained significant attention for their simplicity and intuitive appeal, but also for their incompatibility. This chapter provides a perspective on the meaning and consequences of these and other fairness criteria using graphical models which reveals Equalized Odds and related criteria to be ultimately misleading. An assessment of various graphical models suggests that fairness criteria should ultimately be case-specific and sensitive to the nature of the information the algorithm processes.