Protocol ID: LL-FDE-PART-001
Version: 0.1.0
Status: proposed; not an approved consent form, privacy notice, ethics determination, or authorization for real-data intake

1. Purpose and authority boundary

This protocol defines the minimum participant-rights system for interviews, surveys, work-sample pilots, reference research, and outcome linkage. The accountable ethics/privacy authority must determine what laws, regulations, institutional review, exemptions, or non-applicability decisions govern each study and jurisdiction.

The protocol is informed by OHRP informed-consent guidance and 45 CFR part 46 as applicability inputs, NIST digital-identity privacy principles, and the research-reporting sources in the Source Ledger. It does not assert that HHS rules apply, that an exemption exists, or that any authority has approved this work.

Consent requires understandable disclosure, a voluntary affirmative choice, ability to ask questions, a retained record of the version shown, and a usable withdrawal/correction route. Silence, platform terms, employment, payment, training enrollment, public-profile availability, or an earlier unrelated interview is not study consent.

Separate choices must be recorded for:

participation;
audio/video recording;
deidentified analysis;
aggregate publication;
attributed quotation;
future contact;
future related use; and
future unrelated use.

Bundling optional uses into participation is prohibited unless the accountable authority documents why separation is impossible and permissible.

3. Required participant disclosure

Before collection, provide:

study identity, sponsor, responsible researcher, and verified contact routes;
purpose, activities, time, tools, location/mode, and what is experimental;
why the person was contacted and eligibility basis;
foreseeable risks, discomforts, confidentiality limits, and benefits without exaggeration;
data collected, sources, recording, inference, AI-assisted processing, and prohibited inputs;
who may access which data and whether cross-border/provider transfer occurs;
public, controlled, metadata-only, and unavailable release plans;
retention, deletion, deidentification, backups, legal holds, and audit-trail limits;
incentive and treatment after partial participation or withdrawal;
voluntary participation and non-retaliation/no-effect statement;
accessibility, language, accommodation, and alternate-mode options;
withdrawal, correction, access, complaint, incident, and emergency routes;
whether results, individual feedback, employer reports, or credentials will exist; and
the ethics/privacy determination and responsible authority.

This copy is a drafting frame, not approved participant text:

You are being invited to [study activity] about [plain-language purpose]. LockedIn Labs Research
is the sponsor/conductor. Participation is voluntary. Choosing not to participate or stopping will
not affect employment, an application, training, platform access, customer work, payment already
earned, or a benchmark result.

If you participate, you will [activities] for about [burden]. We will collect [minimum data].
[Recording] is optional / required for [reason] and has a separate choice. Your information may be
seen by [roles] for [purposes]. We plan to publish [aggregate/deidentified material], but we will not
publish [identity/customer/confidential boundaries]. The main risks are [study-specific risks and
mitigations].

You may skip optional questions, request an accessible format, pause, stop, correct information,
or withdraw according to [operational route and limits]. Questions: [monitored route]. Concerns:
[independent/private route]. Data rights and retention: [approved summary/link].

Do you consent to participate under consent version [ID/version/digest]?

Every bracketed field must be completed and reviewed. Do not field placeholder text.

Store the signed/affirmed receipt separately from research data. It records:

study ID, participant key, consent policy ID/version/digest, language, format, and timestamp;
person/process obtaining consent and mode;
each granular choice;
questions or comprehension support provided without storing unnecessary sensitive content;
withdrawal/correction route provided;
legal/ethics determination binding; and
supersession or reconsent history.

The public repository receives neither the receipt nor identity. Participant-level study records store only the permitted state and receipt digest.

6. Comprehension and accessibility

Test consent content cognitively before launch with people outside the drafting team.
Provide screen-reader-compatible, keyboard-accessible, zoom/reflow-safe, captioned/transcribed, and plain-language versions appropriate to the modes.
Offer a human explanation and sufficient time; do not rely exclusively on timed modal text.
Record language and version, not a subjective “English proficiency” judgment.
An accommodation request is not evidence of capability and must remain outside scoring and eligibility unless directly necessary and lawfully reviewed.
When translated versions differ materially, pause that language until reconciled; do not improvise legal/rights language in session.

7. Data separation and permitted use

Use separate stores and role-based access for:

contact/identity and consent;
eligibility and sampling context;
research responses/artifacts;
coding/rating;
protected attributes collected for approved fairness analysis;
outcome records; and
public aggregates.

Study keys are scoped and non-semantic. Do not place names, email addresses, phone numbers, identity documents, employers/customers where identifying, secrets, raw references, or direct identifiers in public schemas, logs, URLs, analytics, screenshots, prompts, or repositories.

AI providers receive participant material only when the approved protocol, consent, provider terms, security/privacy assessment, data location, retention, access, and human verification explicitly allow it. “Deidentified” is not asserted merely because a name was removed.

8. Withdrawal and correction

The participant route must authenticate the request proportionately without demanding unnecessary identity documents. Record:

request time and scope;
identity/authority verification method;
data locations and processing stages affected;
what can be stopped, corrected, deleted, or excluded;
limits such as already published aggregates, backups, or required audit/legal records;
downstream notifications and analysis impact;
completion decision and date; and
appeal/escalation path.

Withdrawal does not require a reason. Payment treatment follows the disclosed incentive policy. A withdrawn record cannot remain count-eligible.

9. References and outcome reports

References and employers are separate participants when the study collects information from them. Candidate consent does not replace reference consent, and employer authority does not replace practitioner consent. Outcome linkage requires separate, specific disclosure of data source, period, fields, access, possible consequences, and withdrawal limitations.

No employer receives a participant’s research response, score, comparison, or status unless a later protocol expressly authorizes it with a lawful basis, valid consent or other authority, contestability, and measurement/legal review.

10. Incidents and retaliation

Immediately hold affected intake for lost consent evidence, unauthorized access/use, confidentiality breach, coercion, employer/sponsor interference, retaliation concern, inaccessible consent, or material mismatch between disclosure and processing. Use the security and participant-conduct channels; do not force a person to complain through their employer or study interviewer.

11. Activation checklist

Applicable authority issued a documented ethics/human-participant determination.
Privacy, security, legal, accessibility, language, data-flow, retention, and provider reviews passed.
Consent and notice content are complete, versioned, cognitively tested, and approved.
Receipt, withdrawal, correction, complaint, incident, and participant-support routes are live and tested.
Data stores, permissions, audit, deletion, backup, and identity separation were technically verified.
Recruiters/interviewers completed training and escalation simulation.
Fieldwork receipt authorizes the exact activity and version.

Until every item is evidenced, real-data intake remains blocked.

Participant consent, rights, and protected-data protocol