1 paper across 1 session
We develop hypothesis tests for fairness metrics for small and large subgroups