For each group (i.e race), the False Positive rate is the same for each one.
- Involves creating sub-models for each group that classifies reaching the expected false positive rate
- May not be legal in certain countries as a method for statistics if the group is race Used within CWG
Problems
- Certain outputs like SAT score might be more representative of success for one population like another, so its a measurement problem to see if the groups you pick accurately work with your metric