Producer-owned
Tests, reruns, source checks and documentation performed within the production process.
The Lab currently reports 2–3 external methods reviewers and 2 completed independent reproductions. Institution-level totals are kept separate from release-specific attribution: a reproduction is attached to a named release only where that mapping and supporting record can be disclosed.
Tests, reruns, source checks and documentation performed within the production process.
Source audit, blinded second coding or methods challenge by someone other than the primary producer.
Clean-environment reconstruction with a reviewable result, deviations and discrepancy record.
A research object is used outside the production team. Use alone does not certify correctness.
Current totals are reported at institution level while release-specific public records remain bound to the corresponding research object.
| Measure | Current state | Disclosure rule |
|---|---|---|
| External methods reviewers | 2–3 | Named only where reviewer identity and specific artifact are cleared. |
| Independent reproductions | 2 completed | Release-level designation requires an attributable reproduction record. |
| Unresolved corrections | 0 | Resolved corrections remain historically visible if any are recorded. |
| Source/provenance coverage | 100% of published datasets | Dataset pages link source/provenance records. |
| Code/versioning coverage | 100% of reproducible outputs | Reproducible outputs carry code and version identifiers. |
The Lab may report aggregate completed reviews and reproductions where confidentiality or publication rights prevent full public attribution. That aggregate does not permit a specific dataset or paper to carry an independently reproduced badge unless the corresponding record is attached to that release.
The current unresolved-correction count is zero. Corrections are governed by the correction policy and machine-readable correction register.
Institutional metrics: data/institutional-metrics.json