Proof Recorded programmes
reporting updates itself from governed source records| Programme | Participants | Evidence received | State |
|---|---|---|---|
| Sustainable Farming Programme · year 2 | 2,140 case files | 61% of requirement items · composition below | live |
| Nutrient reduction pilot · year 1 | 312 case files | 58% · tracked against the 2030 pathway target | method variation flagged |
| Soil carbon monitoring network | 486 case files | 74% · LOI method held constant · regional cohorts safe above k=5 | live |
Sustainable Farming Programme · year 2
requirements defined once · answered from existing recordsOutcome-state composition · the anti-spin visual
Measured
1,046 · 49%
Not yet measured
584 · 27%
Incomplete chain
306 · 14%
Missing — named
204 · 10%
n=2,140 expected outcomes · a programme cannot quietly report only its successes — every state stays in the total and opens to its evidence chains.
Method / confounder variation · why results can't always pool
| Variation | Affected | Handling |
|---|---|---|
| Soil carbon: LOI vs DUMAS lab methods | 2 cohorts | shown side by side — the tool refuses false aggregation |
| Yield basis: weighbridge vs yield monitor | 184 records | monitor records admitted only with calibration certificates |
| Sampling window slip (weather) | 31 records | retained, flagged, excluded from time-sensitive strata |
Programme rules: no ranking, compliance score or effectiveness badge; missing and not-yet-measured outcomes never
drop out of totals; drill-down ends at governed sources, never a private farm; "recorded" is never "proven effective".