Open access
Aug 2026
Fairness Evaluation Paradox: How Biased Test Data Masks True Group Fairness Assessment
It is found that a systematic evaluation bias is present across all metrics, so the same models on the same test data can support opposite fairness conclusions and mask the mistreatment of the most disadvantaged groups.
Sašo Karakatič, Ivona Colakovic, Tjaša Heričko
· Mathematics · 0 citations