Job-analysis recruitment and sampling operations protocol

Protocol ID: LL-FDE-JA-REC-001
Version: 0.1.0
Status: proposed; no contact, recruitment, screening, referral, panel purchase, or participant intake is authorized

1. Purpose

This protocol operationalizes recruitment for the prospective FDE job analysis. It is designed to collect qualified perspectives across real operating contexts without converting a large convenience sample, famous employer list, or self-selected title population into a representativeness claim.

It is informed by the SIOP Principles, OPM job-analysis guidance, AAPOR disclosure standards, COREQ/SRQR reporting prompts, and Census questionnaire-testing guidance recorded in the Source Ledger. No conformance or endorsement is claimed.

2. Preconditions

Recruitment cannot begin until a real fieldwork-authorization-receipt binds all of the following:

• prospectively registered protocol and permanent study ID;
• frozen population, sampling frame, sample-size rationale, strata, allocation, replacement, stopping, and nonresponse rules;
• authorized source and lawful/contractual basis for every contact channel;
• approved recruitment, screening, consent, privacy, accessibility, language, incentive, and participant-support materials;
• data minimization, identity separation, access, retention, deletion, incident, withdrawal, and correction controls;
• named research, operations, methods, participant-rights, data, security, accessibility, legal, and independent-review roles; and
• exact start/end window and approved collection activities.

Historical or commissioned contacts cannot be silently enrolled into the prospective study. Recontact requires an approved basis and fresh consent for the proposed use.

3. Population architecture

Freeze a target-population statement using this format:

The study seeks perspectives from [work relationship], who [observable experience basis],
within [recency window], across [declared contexts], for the purpose of [job-analysis use only].
It excludes [predeclared exclusions]. It does not claim to represent [explicit noncoverage].

Use evidence of work, observation, accountability, and context rather than title alone. Planned perspective groups may include:

1. hands-on practitioners performing forward-deployed work;
2. accountable engineering, product, delivery, security, or business leaders who directly observe it;
3. customer/partner stakeholders who can describe accepted outputs and consequences; and
4. adjacent-role practitioners used to study overlap and boundary conditions.

4. Coverage matrix

The sampling plan must decide which dimensions are design strata, monitoring variables, or descriptive context. Candidate dimensions include:

• practitioner / accountable observer / customer / adjacent-role perspective;
• responsibility and autonomy level;
• delivery model (embedded, rotating, partner, consultancy, internal platform, other);
• company/product stage and engagement duration;
• technical emphasis and system criticality;
• regulated/high-consequence versus other settings;
• successful, failed, stopped, reversed, and ambiguous work;
• geography, language, remote/on-site pattern, and market context where relevant;
• organization size and customer concentration; and
• access/opportunity conditions that affect which artifacts a person could produce.

Do not fill every cell mechanically. The methods lead documents which contrasts matter to the research questions, which cells are feasible and ethical, and which gaps constrain interpretation.

5. Sample-size and stopping rationale

No numeric target is supplied by this protocol. Before registration, the methods lead must choose and justify:

• precision or decision need for structured ratings, if used;
• information-power or comparable qualitative rationale;
• minimum coverage for predeclared contrasts;
• independently double-coded volume and agreement-uncertainty need;
• anticipated eligibility, completion, duplicate, and withdrawal patterns;
• feasibility and participant burden; and
• stopping rules that do not depend on favorable findings.

“We interviewed as many people as possible,” saturation asserted without a method, and a marketing target are not sample-size rationales.

6. Sampling-frame register

For each source, record:

Public directories and professional networks are not automatically authorized contact sources. Terms, privacy expectations, jurisdiction, and institutional policy require review.

7. Recruitment message contract

Every mode and language version states, before screening:

• LockedIn Labs Research is conducting a proposed job-analysis study;
• why the person may have been contacted and how the contact information was obtained;
• participation is voluntary and declining has no employment, training, platform, customer, or benchmark consequence;
• the activity, expected burden, incentive, recording request, data use, and confidentiality limits;
• no score, credential, hiring recommendation, or employer report will be produced from participation;
• how to ask questions, request an accessible format, stop contact, or report a concern; and
• that eligibility and consent occur before research collection.

The message must not say “selected expert,” “representative sample,” “validated study,” “official benchmark panel,” or similar wording unless a claim receipt supports it.

8. Screening and eligibility

Screening collects only information needed to apply frozen rules. It separates:

1. perspective/relationship to work;
2. specific responsibility basis and observation window;
3. recency;
4. relevant operating contexts;
5. conflicts and LockedIn relationship;
6. language/accessibility needs; and
7. duplicate/previous participation check through a privacy-preserving study key.

Identity and role verification uses the least intrusive approved method. Employer prestige, public portfolio, degree, title, certification, or self-attestation without sufficient detail cannot be the sole basis.

Eligibility adjudicators must not see protected attributes not required for the decision, and the rationale is recorded with the rule version.

9. Contact and disposition codes

Every sampled contact receives exactly one current disposition, with append-only changes:

• not_contacted;
• invited;
• delivered_unknown;
• unreachable;
• screening_started;
• ineligible with approved reason code;
• eligible_not_consented;
• declined;
• consented_scheduled;
• partial;
• complete_quality_pending;
• complete_admitted;
• duplicate;
• withdrawn;
• removed_for_protocol_or_quality_reason; or
• contact_suppressed.

Refusal reason is optional; do not pressure a person to explain. Contact attempts, spacing, channel switching, referral use, and stop-contact behavior follow the frozen plan.

10. Incentives, referrals, and conflicts

• Incentives compensate time; they do not purchase a favorable answer or waive rights.
• Disclose amount, timing, conditions, payer, taxes where relevant, and treatment of partial/withdrawn participation.
• Recruiters and interviewers cannot receive performance pay tied to eligibility, completion count, or findings.
• Referral chains are capped and monitored so one company or network cannot dominate invisibly.
• Recruiters disclose customer, employer, training, investment, personal, and other relevant relationships.
• Employees, customers, contractors, trainees, or applicants require explicit anti-retaliation and authority-separation review.

11. Monitoring without outcome peeking

Operations may inspect coverage, contact, eligibility, completion, burden, accessibility, missingness, deviations, incentives, and complaint rates. They may not change the sample to improve thematic support for the ontology after inspecting findings.

Material undercoverage, recruitment imbalance, unsafe conditions, consent failures, unusual recruiter patterns, participant complaints, or protocol drift can pause a channel or the study. Changes follow the registered amendment process and disclose what was known at the time.

12. Recruitment release table

The study publication reports, subject to privacy and small-cell protection:

• sponsor and conductor;
• population and frames;
• sampling/selection and recruitment channels;
• field dates, modes, languages, incentive, and burden;
• contacted, screened, eligible, consented, completed, admitted, excluded, withdrawn, duplicate, and held counts;
• disposition definitions, nonresponse, replacement, deviations, and missingness;
• achieved context coverage and material gaps; and
• no broader representativeness claim than the design permits.

Counts must reproduce from frozen metadata and an aggregate claim receipt.

13. Current disposition

All recruitment decisions, numeric targets, sources, personnel, messages, and authorization remain open. This protocol cannot be used as an outreach script until the fieldwork receipt is authorized.