create-a-research-protocol

# create-a-research-protocol Specify the study precisely enough that trained staff can execute it consistently. ## When to use - Use before data collection, experimentation, fieldwork, laboratory work, or a registered analysis. - Do not change approved participant, safety, or outcome procedures without the required amendment. ## Procedure 1. Record title, identifier, version, date, sponsor, owners, approvals, sites, and revision history. 2. State background, objectives, hypotheses, endpoints, scope, and definitions. 3. Define population or materials, eligibility, sampling, allocation, comparison, and stopping rules. 4. Write chronological procedures with timing, equipment, instruments, scripts, parameters, tolerances, and decision branches. 5. Assign roles, training, permissions, escalation routes, and separation of incompatible duties. 6. Specify consent, privacy, safety, incident, withdrawal, and accessibility requirements where applicable. 7. Define identifiers, source records, quality controls, storage, access, retention, sharing, and deletion. 8. Predefine analysis, exclusions, missing data, deviations, and reporting. 9. Pilot the protocol, record ambiguity and burden, revise through controlled approval, and train operators. 10. Log every deviation with cause, impact, response, and authorization. ## Worked example Two sites interpret a sampling instruction differently during a pilot. The protocol is revised to include a decision table, example, and verification checkpoint. The change receives approval before full collection, and the training log confirms each operator can apply the rule. ## Done - A versioned protocol document contains objectives, methods, roles, controls, governance, analysis, deviations, and stopping rules - Pilot, approval, training, and version records verify that the procedure is executable and every operator uses the current version