03 · What You Need to Know
A Pilot Study Is Still a Study
What Is a Pilot Study?
A pilot study is a smaller-scale investigation conducted before, or sometimes as an initial phase of, a larger study. Its purpose is usually to test aspects of how the future research will work rather than to provide a definitive answer to the main substantive research question.
A pilot might examine recruitment rates, questionnaire comprehension, intervention delivery, adherence, data-collection procedures, measurement variability, retention, technical processes, or other practical features of the planned study.
The term is sometimes used loosely, so researchers should state exactly what the pilot is intended to learn rather than relying on the label.
What Is a Feasibility Study?
A feasibility study asks whether and how a proposed study, intervention, or procedure can be carried out successfully in a particular context.
Pilot and feasibility studies overlap substantially, and terminology varies across disciplines. A pilot study is often understood as one form of feasibility work conducted on a smaller scale, while feasibility research can address a broader set of questions about whether a future study is practical.
For ethics purposes, the terminology is less important than the activities participants undergo.
Small Sample Size Does Not Create an Ethics Exemption
One of the most persistent misconceptions is that a study becomes too small to require ethics review.
Research-ethics requirements generally do not begin at participant number 20, 50, or 100. If the first five participants undergo a research intervention, provide sensitive information, or allow identifiable private data to be collected, the relevant ethical issues exist for those five people.
A small sample can reduce the number of people exposed to risk. It does not make each participant's rights or welfare less important.
“Preliminary” Describes the Research Stage, Not the Ethics Status
A pilot may produce preliminary evidence. That does not mean participant protections are preliminary too.
The study can still involve informed consent, recruitment, interventions, privacy, confidentiality, compensation, adverse events, vulnerable participants, biological specimens, or identifiable data. Those issues should be assessed before the pilot begins when the applicable framework requires prospective review.
Preliminary research
Describes where the study sits in the broader research program and the tentative purpose of its findings.
Exempt research
Describes a formal ethics status reached when the study satisfies defined exemption criteria under the applicable framework.
A Pilot Can Expose Participants to the Same Intervention as the Main Study
Some pilot studies test the exact intervention intended for the later trial. Participants may receive a behavioral intervention, use a device, undergo repeated assessments, provide specimens, take medication, or experience another research procedure.
The fact that researchers are primarily measuring feasibility rather than effectiveness does not remove those exposures.
Indeed, early studies can sometimes contain additional uncertainty because investigators are still learning whether procedures are practical, tolerable, or safe enough to implement as intended.
A Questionnaire Pilot Can Still Involve Human Participants
Not every pilot is clinically intensive. Researchers frequently pilot questionnaires or interview guides with a small number of people to determine whether questions are understandable.
If individuals merely help researchers proofread wording in a way that does not constitute research under the applicable framework, the ethics analysis may differ. If researchers systematically collect participants' responses, experiences, or identifiable information as part of a research investigation, however, the activity may fall within human-participant research requirements.
The distinction depends on what the activity is designed to accomplish and what information is being obtained.
Testing Recruitment Can Itself Involve Research Participants
A feasibility study may ask whether researchers can recruit a particular population. That sounds administrative, but actually testing recruitment can involve approaching real people, screening them, collecting eligibility information, and perhaps enrolling them into research procedures.
Once potential participants are contacted for a research study, recruitment and screening raise ethical issues concerning privacy, voluntariness, appropriate information, eligibility data, and sometimes sensitive characteristics.
Researchers should not assume that “we're only testing recruitment” means recruitment can begin before the applicable ethics process is complete.
Testing Consent Procedures Can Require Prior Ethics Review
A pilot may also examine whether potential participants understand a consent form or whether a recruitment script communicates information effectively.
If researchers systematically collect information from individuals to evaluate the consent process as part of human-participant research, the pilot itself may require an ethics determination.
There is an obvious circularity otherwise: researchers cannot simply use unreviewed participants to test whether the consent process for a future reviewed study works.
Method Development Is Not Automatically Human-Participant Research
There is an important boundary. Researchers can conduct preparatory work without necessarily involving human participants.
Programming a survey, testing database logic with synthetic records, rehearsing interview procedures among research-team members, calibrating equipment without collecting research information about people, or simulating recruitment workflows may not constitute human-participant research.
Once real participants, identifiable private information, or human biological materials enter the activity, reassess whether the applicable research-ethics framework is engaged.
Using Real Records to “Test the Analysis” Can Also Trigger Ethics Questions
A researcher may want to test statistical code or data-cleaning procedures before the main study by opening a small subset of real patient, student, or employee records.
The fact that the researcher calls this a technical pilot does not determine its ethics status. If identifiable private information is being obtained or analyzed for research, the same questions concerning secondary use of existing data can arise.
Where possible, synthetic or appropriately nonidentifiable data may allow some technical procedures to be tested without involving real participants' information.
A Pilot Can Qualify for Exemption
Saying that pilots may require ethics oversight does not mean every pilot requires full committee review.
A low-risk pilot involving an anonymous adult survey, ordinary educational activity, benign behavioral procedure, or another eligible method may satisfy an exemption category under the applicable framework. Other pilots may qualify for expedited review.
The review pathway follows the actual protocol, not the word pilot.
Approval for the Pilot Does Not Automatically Approve the Main Study
A pilot and the later definitive study may differ in sample size, sites, interventions, eligibility criteria, measures, duration, or analytic objectives. Ethics approval for one should not be treated as blanket approval for the other unless the ethics body has specifically reviewed the protocol structure that includes both phases.
If the main study was included prospectively in the approved protocol with clear criteria governing progression from pilot to main phase, institutional procedures may permit the project to continue accordingly. Otherwise, an amendment or new submission may be necessary.
Changes After the Pilot Should Be Reflected in the Ethics Documentation
The entire point of piloting is to learn what needs changing. A questionnaire may be rewritten, eligibility criteria adjusted, recruitment expanded, intervention duration changed, or new measures added.
Those modifications can affect participant information, risk, consent, data collection, or review classification. Follow the applicable amendment process before implementing material changes in the main study.
Data From the Pilot May or May Not Belong in the Main Analysis
Researchers sometimes want to include pilot participants in the final dataset. Whether that is methodologically defensible depends on the design and whether procedures changed between phases.
Ethically, the approved consent and protocol should also be considered. Participants should not be told their information is being collected only to test procedures if investigators actually intend to combine their data with the definitive study in ways not covered by the approved plan.
Plan this possibility prospectively when feasible.
A Pilot Is Not a Loophole for Starting the Main Study Early
Calling the first few participants a “pilot sample” does not authorize researchers to begin an otherwise reviewable study while the full protocol is still awaiting ethics approval.
If those individuals undergo the same research procedures contemplated in the pending protocol, they are participants in research regardless of the informal label placed on their enrollment.
Watch Out
Do not recruit the first few participants “just to test things” while ethics approval for the study is still pending. If the pilot itself involves covered human-participant research, obtain the required approval or determination before recruitment and data collection begin.
Pilot Studies Should Have Their Own Scientific Justification
Ethical research should have sufficient value to justify participant involvement. A pilot therefore needs a clear question of its own: what uncertainty about feasibility, procedures, recruitment, measurement, intervention delivery, or study operations is it intended to resolve?
A vague plan to “try the study on a few people and see what happens” provides a weak basis for exposing participants to burden or risk.
Ethics review can appropriately ask whether the pilot is designed well enough to answer the feasibility question that supposedly justifies involving participants.
Do Not Use Pilot Status to Excuse Weak Consent or Data Protection
Preliminary research still generates real records. Audio recordings, questionnaire responses, clinical measurements, biological specimens, contact information, and screening data require appropriate handling from the first participant onward.
Consent documents and privacy protections should accurately describe what the pilot involves, what will happen to participants' information, whether data may be retained for the later study, and what happens if the main study never proceeds.
