Yield Effect Attributability Score (YEAS)

GMI-AGR-002Agriculture and Farm OperationsIntegration-RequiredPattern O-32: Change Attribution Integrity

Definition

The share of the yield or margin difference observed on AI-prescribed zones that can be confirmed as attributable to the prescription after controlling for weather, hybrid, soil, and model version changes, using the counterfactuals counted in PCCR.

The Measurement Gap

There is no metric that tells you what portion of an observed yield change can be confirmed as attributable to the AI prescription rather than to weather, hybrid, soil, or an unlogged model version change, which means prescriptions are credited and blamed on noise and the engine is retuned toward it.

Formula

YEAS = 1 − |Δy_raw − Δy_adj| / |Δy_raw|, where Δy_adj is the prescription effect from a mixed model with fixed effects for weather index, hybrid, soil unit, and model version, and random effect for field; reported with the 90 percent interval on Δy_adj; YEAS is set to 0 when the interval on Δy_adj spans zero.

Proposed formula, pending validation pilots.

At a Glance

Unit
Score (0 to 1)
Trend Direction
Higher
Category
Agronomic Decision Validation

Decision Supported

Whether to credit, blame, or retune the prescription engine on this season's result, and whether a reported "yield lift" is evidence or noise.

← Stratified Recommendation Accuracy Divergence (SRAD)Zone Specification Fidelity Rate (ZSFR) →

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