Evidence methodology · current
Methodology v0.2
Current demonstration method for checkpoint-specific evidence strength, primary endpoints, raw quantitative context, and explicit insufficient-data states.
Published
July 21, 2026. This route is an immutable public method record.
Rule
Baseline and later measurements must identify the measurement type, checkpoint, applicable denominator, exclusions, and missingness. A quantitative threshold is tested against the recorded raw measure, not a satisfaction vote.
Rule
A paired result is displayed only when the record contains same-person outcome data. Cross-sectional checkpoints are never converted into a paired trend or transition table.
Rule
Day-90 eligibility must distinguish not-yet-due, withdrew consent, returned, protocol exclusion, invalid evidence, no longer eligible, eligible/no response, and completed/analysed whenever source data supports those states.
Rule
Evidence strength is endpoint-specific. A label requires a completed sample, stated verification coverage, a valid measurable claim, a predefined primary endpoint, an attached methodology version, documented missing-data handling, and a recorded manipulation-review status. If those fields are unavailable, the record must say so rather than infer qualification.
Rule
Limitations—including source selection, attrition, incomplete measures, unverified use, and unsupported variation statistics—remain attached to the record and cannot be converted into favourable evidence.
Demonstration evidence is fictional and does not constitute a live study, product conclusion, or production operating claim.