Public claim receipt operations

Protocol ID: LL-FDE-CLAIM-OPS-001
Version: 0.1.0
Status: active rule for drafting; no new empirical claim is authorized by this document

1. Rule

Every objective public statement receives one stable claim ID and one current receipt. The receipt preserves requested wording, exact permitted wording, evidence, scope, uncertainty, prohibited inferences, allowed surfaces, expiry, reviewers, authority, and supersession.

The claim system covers the website, publications, papers, press, social content, sales material, API, downloadable reports, citation metadata, and future result/credential surfaces. Changing the channel cannot upgrade the evidence.

2. Claim classes

Class	Example	Minimum evidence path
Status	“The ontology contains 55 released families.”	Reproducible released artifact and manifest
Method	“The experimental scorer is deterministic for the supplied synthetic fixture.”	Code, fixture, environment, replay, limits
Source synthesis	“Reviewed first-party records describe…”	Dated source set, synthesis method, citations, counterevidence/limits
Occurrence/count	“We interviewed N eligible practitioners.”	Aggregate study receipt and eligible-record manifest
Descriptive finding	“Theme X appeared under Y coding rule.”	Authorized study, codebook, denominator, analysis/agreement/uncertainty
Measurement	“Ratings support interpretation Z.”	Intended-use validity evidence and independent review
Comparative	“Group/context A differs from B.”	Preregistered contrast, sampling/context, uncertainty, fairness/privacy review
Predictive	“Evidence predicts outcome Y.”	Prospective outcome study, replication, calibration/error, use review
Legal/conformance	“Compliant, conformant, certified, accredited.”	Governing authority and qualified legal/conformance evidence
Marketing	“Leading, rigorous, trusted, top 5%.”	Evidence matching literal and implied meaning; otherwise prohibited or qualified

The existing aggregate-claim-receipt remains mandatory for study counts/findings. A general publication receipt cannot replace it.

3. Draft intake

CLAIM ID:
REQUESTOR / OWNER:
REQUESTED WORDING:
CLAIM CLASS:
INTENDED AUDIENCE:
SURFACES:
POPULATION / CONTEXT / GEOGRAPHY / TIME BOUNDARY:
IMPLIED MEANING A REASONABLE READER MAY TAKE:
DECISION OR BEHAVIOR THE CLAIM MAY AFFECT:
EVIDENCE REFERENCES AND DIGESTS:
AGGREGATE CLAIM RECEIPT, IF EMPIRICAL:
UNCERTAINTY / MISSINGNESS / ALTERNATIVE EXPLANATIONS:
PROHIBITED INFERENCES:
EXACT PERMITTED WORDING:
STATUS LABEL:
REVIEWERS / APPROVERS / CONFLICTS:
EXPIRY OR REVIEW TRIGGER:

Claims review evaluates literal wording, visual placement, chart construction, CTA, footnotes, surrounding copy, and likely implied meaning. A technically true footnote does not cure a misleading headline.

4. Evidence ladder

Use the minimum state that truthfully applies:

unverified — assertion received; not suitable as fact;
partial — some evidence exists; unresolved gap remains visible;
reproduced — authorized reviewer reproduced the exact bounded statement from frozen evidence;
independently_reviewed — a qualified independent reviewer inspected the evidence and signed the stated scope.

Published, peer reviewed, validated, representative, and certified are not evidence-ladder synonyms.

5. Approval matrix

Every authorized claim requires claims reviewer, publication guarantor, and release authority. Add:

research director and aggregate receipt for occurrence/descriptive findings;
measurement lead and independent methods reviewer for measurement/comparative/predictive claims;
privacy/participant-rights/accessibility approvals where people or subgroup/context data are involved;
legal-use reviewer for employment, compliance, certification, regulated, or comparative advertising claims; and
conflict reviewer when the sponsor, author, service line, training product, or evaluated party benefits materially.

The same actor cannot be the sole evidence producer, claims reviewer, guarantor, and release authority.

6. Exact wording and placement

The receipt authorizes a string and declared surfaces, not an idea. Preserve:

qualifiers next to the claim rather than behind a generic methodology link;
denominator, time, population, and uncertainty next to a number;
proposed, synthetic, public draft, not validated, or other status label where material;
source/date near current-market claims; and
prohibited-use language where a score-like visual could invite overinterpretation.

A headline, chart label, metadata description, or social snippet needs its own receipt or an explicit surface binding.

7. Expiry, monitoring, correction, and supersession

Every claim declares a date or event trigger: source revision, release change, new data, reviewer withdrawal, incident, correction, legal change, or scheduled review. Expiry moves the claim to held until reauthorized.

A receipt is immutable. Changes issue a new receipt linked by supersedes_receipt_id. Correction or retraction follows the Publication Integrity Workflow across every surface.

8. Machine contracts

Study counts and findings: SCHEMAS/research-study/0.1.0/aggregate-claim-receipt.schema.json
All public claims: SCHEMAS/research-ops/0.1.0/publication-claim-receipt.schema.json

Synthetic fixtures are held by design and cannot authorize public empirical wording.

9. Current high-risk claims

Until real receipts exist, do not state or imply:

500 practitioners or 50 executives were interviewed as an established fact;
fewer than 5% of practitioners would pass;
the benchmark predicts performance or hiring success;
a person is verified, certified, job-ready, or among a percentile;
the method is independently reviewed, peer reviewed, validated, representative, accredited, or conformant; or
a named organization partners with, endorses, or uses this benchmark without documentary authority.