Each integrated evidence family is documented with its source, date, geographic level, role and integration status.
BOI RESEARCH · PROJECT 1
Reproducibility
Research infrastructure for transparent reconstruction, review and independent extension of the BOI agricultural supply-network model.
Reproducibility status: analytical workflow documented and reproducible. Empirical validation remains pending.
01 · REPRODUCIBLE WORKFLOW
How the research can be reconstructed
The road-routing layer is based on OpenStreetMap data and uses a documented market-to-market network construction procedure.
Price-gap screening, transport scenarios, robustness analysis, corridor prioritization and normalized flow allocation are defined through explicit computational procedures.
Transport-cost assumptions and scenario results are recorded explicitly so that the analytical baseline can be independently reviewed.
Research packages contain validation checks, manifests, status files and documented limitations.
BOI distinguishes analytical reproducibility from empirical validation and does not present the current baseline as empirically validated.
02 · RESEARCH RECORD
Documented analytical layers
- Market-price evidence and commodity normalization.
- Road-network construction and market snapping.
- Market-to-market distance calculation.
- Price-gap opportunity screening.
- Transport-cost sensitivity analysis.
- Normalized multi-product flow optimization.
- Route robustness and corridor prioritization.
- Spatial scenario analysis and stress testing.
- Validation and quality-control gates.
03 · DATA GOVERNANCE
Evidence is not treated equally
BOI separates public secondary evidence, analytical derivatives and primary empirical observations. Source provenance and geographic level are preserved during integration.
Province-level evidence is not automatically assigned to individual market locations, and missing empirical observations are not replaced with synthetic values.
04 · VALIDATION BOUNDARY
What reproducibility does not mean
Reproducing the calculations does not constitute empirical validation. BOI must still obtain sufficient transaction-level observations, calibrate the relevant parameters and test the model against an independent holdout dataset before making empirical performance claims.
05 · SCIENTIFIC DISCLOSURE
Current research boundary
The analytical baseline is finalized and reproducible. The current evidence base does not justify an empirical validation claim.
Future empirical observations will be integrated through documented validation gates before any recalibration or out-of-sample performance claim is published.