Public methods overview · Version 1.0
Make the evidence-to-action pathway inspectable.
P4L uses this framework to evaluate sources, bound claims, distinguish interpretation, support institutional decisions, govern public versions, disclose material relationships, and correct consequential errors.
Status: This is P4L’s public methods overview. It is not an independently issued Evidence and Trust Standard, a certification program, a clinical-evidence grade, or a substitute for scientific, clinical, legal, regulatory, or reimbursement review.
Workflow
Signal → Evidence → Decision → Adoption → Impact
Controls
What public P4L work should show.
Source hierarchy
- Primary or authoritative records for current institutional, program, regulatory, and policy status
- Peer-reviewed human evidence and authoritative guidance for health and clinical claims
- Secondary sources only when their added interpretation is necessary and attributable
Claim classes
- Source fact: directly supported by the cited record
- P4L interpretation: a bounded synthesis or implication
- Hypothesis: a proposition requiring further evidence
- Scenario: a conditional future possibility, not a forecast
Evidence strength
- Established human evidence
- Emerging human evidence with limitations
- Mechanistic or preclinical evidence
- Unknown, conflicting, unsupported, or overstated
Status precision
Proposed, announced, funded, selected, accepted, enrolled, authorized, cleared, launched, operational, completed, effective, and scaled are not interchangeable.
Outcome precision
A biomarker, digital signal, engagement measure, delivery event, functional outcome, clinical outcome, operational result, and economic result must not be treated as equivalents.
Context and safety
Population, setting, intended use, material limitations, uncertainty, access, equity, burden, safety, and relevant regulatory boundaries travel with the claim.
Review and versioning
Substantive artifacts should identify author, reviewer status, publication date, material update date, version, source-check date, and correction path. No reviewer is implied when none has approved the work.
Disclosure
Material financial, advisory, employment, sponsor, personal, or institutional relationships that could affect interpretation should be disclosed at the artifact level.
Generative AI
AI may assist ideation, structure, editing, metadata, source discovery, coding, and repetitive checks. It is not the authority for consequential factual, health, policy, regulatory, legal, or biographical claims; a human must verify and approve them.
Corrections
A visible pathway, not a disclaimer.
Material errors are corrected on the affected artifact with a dated note when the change affects meaning. Minor typographic or formatting repairs may be made without a correction entry when they do not alter substance.