Module 2.3.2 · Phase 2 Decision · Step 2.3 Bias Identification & PrioritizationPreferred
Feature Analysis & Proxy Detection
Which features stand in for a protected attribute?
Finds features that stand in for a protected attribute, so discrimination through them does not go unnoticed.
What it does
On the ingested dataset, type-aware correlations and a proxy chain search show which features carry that signal. You then record a keep, transform, remove or monitor decision, with a written rationale, for every feature.
Why it matters
This step identifies and classifies concrete bias mechanisms affecting the system and prioritizes them based on impact, severity, and feasibility of mitigation. It focuses attention on the most material sources of unfairness, preventing diffuse or unfocused intervention efforts.
Where it sits
- Fairness Definition
- Populations & Intersections
- Bias Identification & Prioritization
- Metrics & Thresholds
- Bias Detection & Prioritization
- Feature Analysis & Proxy Detection
- Explainability (XAI)
How to use it
In the Fairness Navigator Beta
Work through the steps in order. Several modules open from the Navigator's steps and hand their results back.
On its own Live
Open 2.3.2 from the platform's Modules page and run it when your project needs it. Work is saved with the assessment.
After a Pulse reading Live
Start with a fast first read, then go deeper with the modules where it points.