Research

Findings

Findings, evidence links, and review.

FindingsSelect study

Reference & policy

Review a finding →Release controls

01Draft interpretation

02Evidence review

03Challenge review

04Method review

05Approved finding

06Release candidate

07Released

Finding evidence chain

A finding should be inspectable from claim back to source.

The review record does not merely collect citations. It preserves what supports the interpretation, what challenges it, the conditions under which it appears to hold, and what remains uncertain.

Proposed findingState the interpretation narrowly enough that supporting and challenging evidence can be reviewed.

Supporting evidenceLink the finding to source excerpts, coded segments, structured responses, matrices, or documented observations.

Contrary evidenceRecord credible evidence, cases, or interpretations that weaken, qualify, or complicate the proposed finding.

Context and conditionsIdentify the institutional, historical, group, operational, or situational conditions within which the finding appears to hold.

Limitations and uncertaintyState missing evidence, sampling limits, measurement constraints, alternative explanations, and unresolved uncertainty.

Review decisionRecord who reviewed the finding, what changed, the disposition, and whether it may move toward release.

Challenge review

The strongest interpretation is not necessarily the first coherent one.

Before a finding is approved, the reviewer should actively test whether another explanation fits the evidence, whether important cases do not fit, and whether institutional history, group relationships, operational practices, or missing information change the conclusion.

Required challenge questions

What evidence most directly supports the proposed finding?
What credible evidence weakens, contradicts, or qualifies it?
Which alternative explanation remains plausible?
What contextual condition would make this finding different elsewhere?
What evidence is missing, inaccessible, or too uncertain to interpret confidently?
What would cause the finding to be revised or withdrawn?

Release controls

Publication is a separate decision from analytical approval.

A finding may be analytically useful without being safe, appropriate, or ready for external release. The release layer therefore reviews evidence, quotation use, privacy, method disclosure, versioning, and authority separately.

01Evidence chain completeReleased claims remain traceable to approved evidence and analysis records.

02Quotation reviewQuoted material is checked against consent, confidentiality, contextual integrity, and re-identification risk.

03Disclosure reviewOutputs exclude restricted identity information and are reviewed for indirect identification risk.

04Method disclosureThe release describes methods, scope, material limitations, and analytical boundaries at an appropriate level.

05Human release authorityA designated reviewer approves the release version; machine assistance cannot release findings autonomously.

06Controlled exportExported reports, tables, excerpts, and datasets follow study permissions, access rules, and version controls.

Output boundary

Released output is a governed derivative of the research record.

Reports, briefs, presentations, public articles, contributor materials, data extracts, and client deliverables may present only the portions of the research record appropriate to their audience and permissions. Release never changes the underlying evidence, erases contradictory material, or converts provisional machine output into an approved finding.