MENA Open Data & Evidence Lab
Evidence & validation

Independent review is an institutional asset, not a decorative badge.

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.

External methods reviewers2–3Current institutional total.
Independent reproductions completed2Completed outside reproductions.
Unresolved corrections0Current correction state.
Reproducible outputs with code/versioning100%Current published reproducible-output coverage.
01 / Internal QA

Producer-owned

Tests, reruns, source checks and documentation performed within the production process.

02 / Independent verification

Different person

Source audit, blinded second coding or methods challenge by someone other than the primary producer.

03 / Reproduction

Outside execution

Clean-environment reconstruction with a reviewable result, deviations and discrepancy record.

04 / External use

Actual use

A research object is used outside the production team. Use alone does not certify correctness.

Institutional validation state

Current totals are reported at institution level while release-specific public records remain bound to the corresponding research object.

MeasureCurrent stateDisclosure rule
External methods reviewers2–3Named only where reviewer identity and specific artifact are cleared.
Independent reproductions2 completedRelease-level designation requires an attributable reproduction record.
Unresolved corrections0Resolved corrections remain historically visible if any are recorded.
Source/provenance coverage100% of published datasetsDataset pages link source/provenance records.
Code/versioning coverage100% of reproducible outputsReproducible outputs carry code and version identifiers.

Release-specific versus institution-level evidence

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.

Correction standard

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