methodEnglish

Test causal claims, assumptions and alternatives

Show why the pathway might—or might not—hold in this context.

EnglishEspaƱolFranƧais

Documentation follows the current system language. Missing translations fall back to the page's canonical language.

Evidence and assumptions

Causal-claim review matrix

Do not rely; clarify the link

Evidence limited · Mechanism unclear

Treat as a hypothesis; prioritize learning

Evidence limited · Mechanism plausible

Record the gap and test transferability

Evidence limited · Mechanism reviewed

Map alternatives and uncertainty

Evidence mixed · Mechanism unclear

Use cautiously with explicit assumptions

Evidence mixed · Mechanism plausible

Preserve disagreement and monitor

Evidence mixed · Mechanism reviewed

Check relevance and adverse effects

Evidence stronger · Mechanism unclear

Use as a bounded contribution claim

Evidence stronger · Mechanism plausible

Review, monitor, and remain open to revision

Evidence stronger · Mechanism reviewed

Confidence is a transparent judgment, not a measured probability. Challenging evidence remains visible.

For each material link, distinguish the proposed mechanism from the assumptions required for it to operate. Add supporting and challenging evidence with provenance, relevance and limits. Confidence is a documented human judgment, not a statistical probability or a substitute for evidence.

Ask what else could produce the outcome, where evidence comes from, whether it transfers to this population and context, and what would change the group’s mind. Record contradictory findings and unresolved disagreement. Do not turn expert opinion, participant testimony or an assistant suggestion into a verified fact.

Critical assumptions deserve a validation or learning action, owner, timing and trigger. The Risk & Assumption Stress Test can challenge selected claims in greater depth; returning work should retain its source model and version so the pathway remains traceable.