SCC Nexus Methodology

Research lifecycle

SCC Nexus uses a staged lifecycle intended to prevent discovery artefacts from becoming public claims without adequate scrutiny.

1. Question definition

Define the research question, scope, population or system of interest, timeframe, decision context and intended audience. Record material assumptions before analysis where feasible.

2. Source discovery

Search authoritative databases, official repositories, scholarly indexes and other relevant sources using documented search terms and dates. Discovery results are candidates, not evidence conclusions.

3. Source screening

Assess relevance, source identity, duplication, currency, licensing and minimum evidential suitability. Record exclusions when they could materially affect interpretation.

4. Extraction

Capture source identifiers, dates, variables, study or record characteristics, extraction method and provenance. Preserve original units and fields where practicable.

5. Corroboration

Cross-check important facts against independent or primary sources where available. Resolve entity identity deterministically using stable identifiers before relying on fuzzy matching.

6. Analysis

Use deterministic, versioned transformations where possible. Separate descriptive analysis from inferential analysis and record assumptions, thresholds, models and sensitivity checks.

7. Validation

Run schema, completeness, consistency, range, duplication and provenance checks appropriate to the dataset. Scientific interpretation receives human review where material.

8. Publication

Publish only outputs that pass the defined gate. Every public research output should include citations, method version, output version, limitations and status.

9. Maintenance

Re-run research according to a declared cadence where data are time-sensitive. Record corrections, retractions, source changes and methodological changes.

Review states

discovered — identified during search.

screened — relevance and source identity checked.

verified — evidential support and provenance checked under the applicable protocol.

published — passed the public-release gate.

These states are not interchangeable. A source can be authentic yet fail to support a claim.

Minimum publication gate

A substantive public output should fail closed if any mandatory item is absent: source provenance, method version, validation status, limitations, output timestamp/version, or required reviewer decision.

Source version: METHODOLOGY.md. Public HTML is the canonical presentation page.