Fairness Audits of Institutional Risk Models in Deployed ML Pipelines

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: McConvey, Kelly, Das, Dipto, Ghai, Maya, Zhai, Angelina, Lee, Rosa, Guha, Shion
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917426049843200
author McConvey, Kelly
Das, Dipto
Ghai, Maya
Zhai, Angelina
Lee, Rosa
Guha, Shion
author_facet McConvey, Kelly
Das, Dipto
Ghai, Maya
Zhai, Angelina
Lee, Rosa
Guha, Shion
contents Fairness audits of institutional risk models are critical for understanding how deployed machine learning pipelines allocate resources. Drawing on multi-year collaboration with Centennial College, where our prior ethnographic work introduced the ASP-HEI Cycle, we present a replica-based audit of a deployed Early Warning System (EWS), replicating its model using institutional training data and design specifications. We evaluate disparities by gender, age, and residency status across the full pipeline (training data, model predictions, and post-processing) using standard fairness metrics. Our audit reveals systematic misallocation: younger, male, and international students are disproportionately flagged for support, even when many ultimately succeed, while older and female students with comparable dropout risk are under-identified. Post-processing amplifies these disparities by collapsing heterogeneous probabilities into percentile-based risk tiers. This work provides a replicable methodology for auditing institutional ML systems and shows how disparities emerge and compound across stages, highlighting the importance of evaluating construct validity alongside statistical fairness. It contributes one empirical thread to a broader program investigating algorithms, student data, and power in higher education.
format Preprint
id arxiv_https___arxiv_org_abs_2604_19468
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Fairness Audits of Institutional Risk Models in Deployed ML Pipelines
McConvey, Kelly
Das, Dipto
Ghai, Maya
Zhai, Angelina
Lee, Rosa
Guha, Shion
Computers and Society
Artificial Intelligence
Human-Computer Interaction
Fairness audits of institutional risk models are critical for understanding how deployed machine learning pipelines allocate resources. Drawing on multi-year collaboration with Centennial College, where our prior ethnographic work introduced the ASP-HEI Cycle, we present a replica-based audit of a deployed Early Warning System (EWS), replicating its model using institutional training data and design specifications. We evaluate disparities by gender, age, and residency status across the full pipeline (training data, model predictions, and post-processing) using standard fairness metrics. Our audit reveals systematic misallocation: younger, male, and international students are disproportionately flagged for support, even when many ultimately succeed, while older and female students with comparable dropout risk are under-identified. Post-processing amplifies these disparities by collapsing heterogeneous probabilities into percentile-based risk tiers. This work provides a replicable methodology for auditing institutional ML systems and shows how disparities emerge and compound across stages, highlighting the importance of evaluating construct validity alongside statistical fairness. It contributes one empirical thread to a broader program investigating algorithms, student data, and power in higher education.
title Fairness Audits of Institutional Risk Models in Deployed ML Pipelines
topic Computers and Society
Artificial Intelligence
Human-Computer Interaction
url https://arxiv.org/abs/2604.19468