Matchings Under Biased and Correlated Evaluations
Amit Kumar, Nisheeth K. Vishnoi · 2025
Read Matchings Under Biased and Correlated Evaluations by Amit Kumar, Nisheeth K. Vishnoi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-23628.txt.
Description
Amit Kumar and a co-author study fairness in matching markets. Two institutions rank candidates from two groups using signals that are biased and only partly aligned. Think shared test scores or AI assessments that each school processes on its own. One parameter measures the bias against the weaker group; another measures how closely the two rankings line up.
Earlier work assumed both schools judge in the same way, and found the chosen share of the weaker group tracks the bias in direct proportion. The new study focuses on markets where all candidates prefer the same school, and the story sharpens. The chosen share reacts in a curved way to how aligned the schools are: small drops in alignment can cause steep falls. The authors find threshold values where selection behavior flips. Their formulas show when better alignment beats bias cuts, and when the reverse holds, for fairer selection design.
Catalog record
| CREATOR | Amit Kumar, Nisheeth K. Vishnoi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 238 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-23628