Demographic Identification Accuracy Differential (DIAD)

GMI-LE-003Law EnforcementResearch-ForwardPattern O-22: Fairness Drift

Definition

The spread in identification accuracy across demographic or vehicle population segments, measured on the agency's own operational output rather than on a vendor benchmark.

The Measurement Gap

There is no metric that tells you how identification accuracy is distributed across population segments in operational use, which means a threshold defensible in aggregate can concentrate its error on one subpopulation without any figure moving.

Formula

DIAD = max_g(A_g) - min_g(A_g), reported with sigma(A_g) across segments g.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage points
Trend Direction
Lower
Category
Identification and Attribution Integrity

Decision Supported

Whether a single agency-wide match threshold is defensible, or whether the system's error is concentrated in a subpopulation.

← Correlated Signal Independence Ratio (CSIR)Hotlist Record Staleness at Hit (HRSH) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment