Pilot Study in Research: Definition, Types, Steps & Examples

A pilot study is a small-scale trial run of a planned larger study, conducted to assess whether the full study is feasible and how it should be designed. It is a rehearsal, not a performance. The findings are not meant to confirm a hypothesis but to tell the researcher whether the plan is workable.

Table of Contents

What Are the Different Types of Pilot Studies?

There are four main types. Choosing the right one depends on the stage of research, the level of certainty about the protocol, and whether pilot participants’ data can realistically be included in the main trial.

Type	Design	Key Feature	Best Used When
External pilot	Stand-alone study before the main trial	Pilot data not included in the main trial analysis	Protocol is untested; changes are expected
Internal pilot	First phase embedded within the main trial	Pilot data retained in final analysis if no major changes needed	Protocol is well-developed; changes unlikely
Randomized pilot trial	Uses randomization on a small scale	Tests allocation, blinding, and adherence for a future RCT	Planning a randomized controlled trial
Non-randomized feasibility study	No random allocation	Tests recruitment, acceptability, and delivery procedures	Early-stage feasibility testing; pre-RCT work

These categories are important because they determine what conclusions can be drawn from pilot data and how that data may be used when creating the main study protocol.

How Is a Pilot Study Different from Similar Studies?

Three study types are often confused with pilot studies. The differences matter for how findings can be interpreted and used:

Study Type	Purpose	Key Distinction from Pilot
Feasibility study	Asks whether a larger study can be done at all	Broader scope; does not necessarily mirror the main study design
Exploratory study	Investigates a phenomenon without aiming for conclusive results	No intent to scale up; not a precursor to a specific main trial
Proof of concept study	Demonstrates whether a concept or technology is workable	Product- or process-focused; narrower aim than a pilot study

Objectives of a Pilot Study in Research

A well-designed pilot study addresses five core objectives, all oriented around feasibility rather than hypothesis testing:

Objective	What It Involves
Assess feasibility	Evaluate participant recruitment rates, retention rates, refusal rates, and whether the numbers of participants, researchers, and support staff are adequate.
Identify resource requirements	Determine how long the intervention or questionnaire takes to administer, whether software and equipment function as expected, and what monitoring procedures need to be in place.
Understand procedural challenges	Surface problems in study design, data collection processes, equipment handling, and logistics before they occur at full scale.
Refine data collection instruments	Verify that questionnaires are relevant, clearly worded, and coherently presented; confirm that outcome measures are valid for the target population.
Develop good clinical and research practices	Establish documentation procedures, informed consent processes, data collection tools, regulatory reporting workflows, and monitoring systems that will carry through to the main study.

How to Conduct a Pilot Study: A Step-by-Step Guide

The methodology for a pilot study mirrors the steps of a full-scale study, but on a smaller, exploratory scale. Each step should be documented in a formal pilot protocol.

Step	Action	Key Considerations
1	Define objectives	Specify feasibility goals and set pre-specified, quantitative benchmarks (e.g., at least 70% of participants will attend 8 of 12 sessions). Vague objectives produce uninterpretable results.
2	Design the study	Select a small, representative sample; develop research instruments such as surveys or interview guides; choose whether to randomize.
3	Obtain ethical approval	Secure clearance from the relevant Institutional Review Board (IRB) or Research Ethics Committee (REC) and obtain informed consent from all participants. This step is mandatory, not optional.
4	Recruit participants	Use eligibility criteria and sampling strategies that match the main study. The sample should reflect the target population.
5	Collect data	Administer the intervention or instrument as planned. Document deviations. Collect data on feasibility, acceptability, and appropriateness of procedures and not just on outcomes.
6	Analyze data	Use descriptive statistics and qualitative feedback. No inferential statistics or hypothesis tests are appropriate. No power analysis is required or recommended.
7	Report results	Produce a detailed report covering key findings, protocol changes made during the process, feasibility benchmark outcomes, and instruments used.
8	Make a go/modify/stop decision	Use pre-specified progression criteria to decide whether to proceed as planned, modify the main study design, or stop and redesign.

What Should Happen After the Pilot Test?

Post-pilot actions are as important as the pilot itself. A pilot that does not lead to explicit design decisions has limited value.

Locate and document challenges: List every issue or risk identified: recruitment shortfalls, dropout patterns, protocol deviations, equipment failures, or participant burden. Develop a backup plan for each.
Determine feasibility: Review whether the pilot achieved its pre-specified benchmarks. If not, the study design should be reconsidered before proceeding.
Refine methods and procedures: Modify data collection instruments such as questionnaires, eligibility criteria, randomization procedures, and outcome measures as needed.
Prepare a detailed main study plan: Update the full protocol to reflect all changes, including revised timelines, adjusted sample sizes, and corrected consent procedures.
Keep pilot data separate: Pilot participants’ data should be reported separately and not mixed into the main study results unless the design explicitly incorporates an internal pilot with pre-specified criteria for data inclusion.

Advantages and Disadvantages of Pilot Testing

Advantages	Disadvantages
Identifies flaws in study design, data collection, and instruments before full-scale investment	Requires additional funding, time, and personnel that may divert resources from the main study
Allows researchers to refine research questions, eligibility criteria, and randomization procedures	Results from small samples are not generalizable to the broader target population
Provides practical insights into participant recruitment and retention strategies	Risk of contamination if pilot participants are included in the main study since they have already been exposed to the intervention
Helps ensure the research is ethically sound by surfacing consent and burden issues early	A successful pilot does not guarantee the success of the main study
Builds funder, ethics committee, and stakeholder confidence by demonstrating empirical feasibility evidence	Pilot studies lack the statistical power to detect treatment effects or produce reliable effect size estimates
Supports development of informed consent procedures, data tools, and monitoring workflows	Problems identified in the pilot are sometimes not acted upon before the main trial begins

How Many Participants Do You Need for a Pilot Study?

There is no universal rule. Sample size in a pilot study should be based on practical considerations, not on power calculations. Because pilot studies are not designed to test hypotheses, a power analysis is neither required nor appropriate.

Factor	Guidance
Participant flow	How many people need to be screened to enroll enough participants? Factor in expected refusal and dropout rates.
Budgetary constraints	What is the maximum number of participants the available resources can support?
Feasibility benchmark requirements	How many participants are needed to generate a meaningful estimate of recruitment rate, adherence, or dropout? A minimum of 10 is often cited for questionnaire pre-testing; 20–30 is common for intervention-based feasibility studies.
Study type	Internal pilots embedded in a main trial may use fewer dedicated pilot participants than stand-alone external pilots.

A common rule of thumb cited in the literature is 10% of the planned main study sample, but this should always be evaluated against the specific feasibility objectives. The goal is to have enough participants to meaningfully evaluate the feasibility benchmarks, not to achieve statistical significance.

What Are the Common Misuses of Pilot Studies?

Pilot studies are frequently misused in ways that produce unreliable findings and contribute to the failure of subsequent main trials. Awareness of these pitfalls is as important as knowing how to design a good pilot.

Misuse	The Problem	The Correct Approach
Testing preliminary efficacy	Pilot is underpowered; conclusions are premature and uninterpretable	Focus exclusively on feasibility benchmarks
Using pilot effect sizes for power calculations	Estimates are unstable; leads to over- or underpowered main trials	Base power calculations on clinically meaningful differences
Assessing safety/tolerability	Sample too small to detect all but extreme adverse events	Report adverse event rates with confidence intervals; do not conclude ‘safe’
Ignoring feasibility signals	Poor recruitment or adherence carries over into a failed main trial	Act on every flagged problem before launching the main study
Mislabelling small studies as pilots	Obscures the true purpose; produces findings of limited utility	Only use the term ‘pilot’ for studies explicitly designed to test feasibility

Ethics and Informed Consent in Pilot Studies

Ethical approval is a mandatory step in pilot studies, not an optional one. Because pilot studies involve human participants, all standard research ethics requirements apply.

IRB/REC approval: Obtain formal clearance from the relevant Institutional Review Board or Research Ethics Committee before any data collection begins.
Informed consent: All participants must receive a full explanation of the study and provide written consent. The consent document for a pilot should reflect the study’s limited scope if activities differ from the main trial.
Participant burden: One of the most valuable outputs of a pilot is an assessment of how burdensome the procedures are. If participants find assessments too long or intrusive, this is a critical finding that must be addressed.
Ethical benefit of piloting: Running a pilot on a small scale before exposing larger numbers of participants to a potentially flawed or unworkable design is itself an ethical act because it protects future participants from unnecessary burden and risk.
Incentives: Determine appropriate participant compensation during the pilot, as this informs the main study budget and helps assess participation motivation.

Challenges of Pilot Testing in Research

Despite their value, pilot studies face practical and methodological challenges that researchers should anticipate:

Resource diversion: Pilot studies require time, funding, and personnel. In resource-constrained settings, these demands may compete directly with the main study.
Limited generalizability: The small sample means that feasibility results may not reflect what will happen in a larger, more diverse population.
Contamination risk: If pilot participants are included in the main study, they may respond differently because they have already been exposed to the intervention or measurement tools.
Sunk-cost pressure: When a pilot requires significant investment, research teams may find it difficult to stop and redesign after an unsuccessful outcome.
Reporting deficiencies: Many published pilot studies do not clearly state feasibility outcomes, pre-specified progression criteria, or how findings changed the main study design. This limits their usefulness to the broader research community.
Misinterpretation: Researchers sometimes interpret a pilot that ‘went well’ as evidence that the intervention works but this is a conclusion the data cannot support.

How to Conduct a Pilot Test of a Questionnaire?

Pilot testing a questionnaire checks whether questions are clear, relevant, and produce usable data before deployment at scale.

Step	Action
Define what you’re testing	Decide whether you’re checking wording clarity, question order, response options, completion time, or the validity and reliability of an existing scale.
Select a pilot sample	Choose 10–30 people who reflect the target population’s demographics, language level, and familiarity with the topic. A minimum of 10 is often cited as sufficient for basic pre-testing.
Administer under realistic conditions	Use the same delivery mode (online, paper, phone, in-person) and setting planned for the main study.
Time the completion	Record how long it takes participants to finish, and flag any sections where people slow down, skip items, or appear confused.
Collect feedback on each item	Ask participants to flag confusing wording, ambiguous response scales, missing options (e.g., “prefer not to say”), or culturally inappropriate questions.
Check data patterns	Look for items with no variation in responses (everyone picks the same option), high rates of “don’t know,” or items left blank.
Test reliability if using a scale	Calculate internal consistency (e.g., Cronbach’s alpha) if the questionnaire includes a multi-item scale being adapted or translated.
Revise and re-test if needed	Make changes based on findings. If changes are substantial, consider a second small pilot before full deployment.

What to look for in pilot responses:

Items consistently skipped or left blank
Response options that don’t cover real answers (forcing “other”)
Questions interpreted differently than intended
Double-barrelled questions (asking two things at once)
Leading or loaded wording
Completion time significantly longer than expected
Technical issues with the survey platform (for online questionnaires)

Techniques for pilot-testing a questionnaire:

Cognitive interviewing: Ask participants to think aloud as they answer, explaining how they interpreted each question
Think-aloud protocol: Similar to cognitive interviewing, used to surface comprehension issues in real time
Debriefing questions: Add a short set of questions at the end asking specifically about clarity, length, and difficulty
Test-retest: Administer the questionnaire twice to the same small group to check for consistency over time

How to Conduct a Pilot Study for Qualitative Research?

Pilot studies in qualitative research test interview guides, focus group protocols, facilitation approaches, and the overall feasibility of generating rich, usable data.

Step	Action
Develop a draft guide	Create the interview or focus group guide with open-ended questions, probes, and prompts.
Recruit a small sample	Select 3–8 participants for interviews, or 1–2 focus groups, who reflect the characteristics of the intended sample.
Conduct in realistic conditions	Use the same setting, format (in-person, video, phone), and recording method planned for the main study.
Reflect on facilitation	Note where questions felt awkward, where participants seemed confused, and where probes were needed but missing.
Review the data generated	Transcribe and review pilot transcripts to check whether the questions produce data relevant to the research aims.
Assess practical logistics	Evaluate session length, recording quality, transcription time, and participant comfort.
Revise the guide	Remove or reword questions that didn’t generate useful responses; add probes where participants needed more prompting.

What to Check When Piloting Interviews

Do questions elicit detailed, reflective answers or short, closed responses?
Is the question order logical and does it build rapport before moving to sensitive topics?
Are there unnecessary or redundant questions that can be cut?
Does the interview run within the planned time?
Are recording equipment and consent procedures working smoothly?
Did any questions cause discomfort, confusion, or distress that needs to be addressed?

What to Check When Piloting Focus Groups

Does the facilitator’s guide allow enough time for group discussion versus individual responses?
Do participants engage with each other, or does the facilitator have to do most of the talking?
Are there dominant voices that need to be managed through facilitation technique?
Does the group size (typically 6–10) work well for the venue and topic?
Are ground rules (confidentiality, respectful disagreement) clearly communicated and effective?
Does the topic guide need reordering based on how naturally the conversation flowed?

This article was originally published on March 13, 2025, and updated on June 12, 2026.