ACADEMIC WRITING GUIDE

How to Conduct and Report a Pilot Study for Your Dissertation (With Worked Examples)

By Acadelyra Editorial TeamPublished

You have designed a questionnaire, written an interview guide, or arranged access to records. Before relying on those procedures for your dissertation, how will you know whether people understand the questions, recruitment actually works, and the information you collect can answer your research question? A well-planned pilot can identify these problems while there is still time to correct them.

A pilot study is not a miniature dissertation with an obligation to produce significant results. Its central purpose is to rehearse and assess the feasibility of selected methods or procedures. This guide explains when a pilot is useful, how to define measurable checks, what to do with the results, and how to write the pilot section honestly. The examples are hypothetical teaching scenarios, not data collected from actual participants.

Quick answer: what is a pilot study?

A pilot study is a small-scale rehearsal or feasibility assessment of aspects of a planned larger investigation. It may test recruitment, consent, survey wording, interview prompts, recording, timing, data handling, or the practicality of an analysis workflow. What it tests depends on the main study. A pilot should have a specific decision attached: proceed as planned, revise a procedure, repeat a focused check, or reconsider feasibility.

Some institutions use “pilot,” “feasibility study,” “pretest,” and “field test” differently. A questionnaire pretest examines question understanding or instrument function; a pilot may rehearse more of the research pathway. Use your department’s terminology and distinguish these purposes clearly.

Do you need a pilot study in every dissertation?

No. A pilot is particularly useful when a procedure is new to your setting, recruitment is uncertain, an instrument has been adapted, interviews may be sensitive, or the logistics involve several steps. It may add little when you analyse an existing, well-documented public dataset and no new collection process needs rehearsal. Some institutions mandate pilots for certain designs; others do not. Ask which uncertainty the pilot will reduce before deciding to conduct one.

For example, the decision to pilot depends partly on how your data collection procedure is organised, including access, consent and the timing of each activity.

Pilot study, questionnaire pretest and main study: the difference

Differences between a questionnaire pretest, pilot or feasibility study, and main study
ActivityPrimary questionTypical output
Questionnaire pretestDo respondents interpret items and response options as intended?Reworded items; improved instructions or routing
Pilot or feasibility studyCan the selected study processes work within real constraints?Feasibility indicators, process changes, go/revise decision
Main studyWhat do the collected data show about the research question?Substantive analysis and appropriately qualified findings

A survey pilot can include a pretest, but the two are not synonyms. For example, five cognitive interviews might identify an unclear word, whereas a rehearsal invitation to 40 eligible students might assess whether the recruitment plan generates enough participants.

How to conduct a pilot study: nine practical steps

1. Define the uncertainty before you collect anything

Start with a practical question, not “I need a pilot because dissertations have one.” Examples include: Can eligible students be reached through an approved recruitment channel? Do interview questions elicit concrete experiences? Does a two-part questionnaire take less than the promised 15 minutes? Can the research team securely extract records without missing necessary fields? Each objective should lead to an observable indicator.

2. Choose the procedure to rehearse and its boundaries

Specify what will be tested: invitation and eligibility screening, consent, the full questionnaire, an interview recording process, a record-extraction form, or the sequence from data capture to clean export. Decide whether the pilot will imitate the whole workflow or only a risky part. A smaller but targeted rehearsal is often more informative than a poorly specified miniature study.

3. Obtain the permissions and participant protections required

Do not assume that calling data collection a “pilot” removes ethics obligations. Depending on your setting, participant recruitment, identifiable records, audio recordings, sensitive questions or a research intervention may require ethics review, site authorisation and informed consent. Approval requirements vary by jurisdiction and institution. The pilot should use the permitted recruitment and data-handling procedures; do not retroactively treat unapproved data as authorised.

4. Set feasible sample and recruitment criteria

Select participants or records that can reveal whether the intended procedure works. The pilot sample should be justified by its feasibility objective, not by a universal rule that every pilot needs 10, 20 or 30 people. If the pilot estimates an invitation response proportion, a tiny sample gives imprecise information. If it explores whether interview prompts are understandable, a small purposive set may be more useful. Report recruitment sources, inclusion criteria, refusals and relevant differences between pilot and target participants.

The rationale should remain consistent with the sampling method and sample size guide, but pilot size is justified by the feasibility decision rather than an assumed main-study effect size.

5. Set decision criteria before seeing the pilot results

Decide what evidence will count as workable. Possible indicators are: the proportion of invitations yielding eligible consents, median completion time, proportion of unusable records, frequency of unanswered items, proportion of recordings intelligible enough for transcription, or whether a planned data export retains all necessary variables. Thresholds should follow actual constraints, relevant guidance or a justified local judgement. Avoid inventing universal cutoffs.

Illustrative pilot-study feasibility measures and decision rules
Feasibility objectiveIllustrative measureIllustrative decision rule
Recruit participantsEligible consents / valid invitationsIf fewer than 1 in 5 invitations yield consent, reconsider recruitment volume or channels
Reduce questionnaire burdenMedian minutes to completeIf median exceeds 15 minutes, shorten or simplify before full launch
Obtain usable survey responsesForms meeting predefined core-item rule / completed formsIf under 90% usable, inspect routing or item clarity
Record interview audio reliablyAudible recordings / consented interviewsIf any critical recording failure occurs, correct equipment and retest

These thresholds are hypothetical planning choices for teaching. Your project may need different criteria. A useful threshold must lead to a real action.

6. Collect process information, not just answers

Capture invitations sent, eligibility failures, consent decisions, interruptions, duration, question-level nonresponse, technical faults and participant comments on wording. Record procedural departures. A participant who says an item was confusing may teach you more about measurement quality than a pilot average score. Collect only information needed and permitted under your approved plan.

7. Review data quality and feasibility against the criteria

Summarise process results transparently. For instance, 8 consents out of 40 valid invitations represent a 20% consent yield; 18 usable returns out of 20 submitted forms represent 90% usability. State the denominator and avoid reporting “90% success” without explaining what was counted. When a criterion is missed, investigate the cause rather than quietly lowering the criterion afterward.

Before any pilot summary, use a consistent research-data cleaning process to distinguish genuine omissions from export or coding mistakes.

8. Document revisions and decide whether to retest

Keep a revision record: original wording or procedure, observed problem, change made, reason for change and whether another check is needed. Revisions might include moving a difficult item, changing response options, clarifying an eligibility screen, adding an interview prompt or improving a secure export. Substantial changes to a tool or protocol may need fresh ethics approval or amendment before use. If a key risk remains untested after revision, rehearse that part again.

9. Explain exactly what the pilot supports—and what it does not

A successful pilot supports confidence that the tested process is feasible under the pilot conditions. It does not establish that the substantive hypothesis is true, that a small observed group difference will recur, or that the entire main study is free of risk. The NIH NCCIH cautions specifically against using underpowered pilot outcomes as preliminary efficacy tests. Apply that principle according to the design and purpose of your own study.

Worked example 1: Quantitative student questionnaire pilot

Proposed main study (hypothetical). A university student plans a cross-sectional survey on the relationship between perceived feedback usefulness and academic engagement among second-year undergraduates. The pilot will test the invitation process, completion time, routing and core-item usability, not that relationship.

Pilot design. Forty eligible students receive an authorised invitation. Participants provide appropriate consent before completing an online questionnaire. The investigator logs response duration, incomplete items, routing faults and clarification requests. Illustrative criteria are consent yield of at least 20%, median duration no greater than 15 minutes, and at least 90% usable completed forms.

Illustrative results and decisions from a quantitative student questionnaire pilot
Pilot observationIllustrative resultConsequence
Consents / valid invitations10 / 40 = 25%Recruitment criterion met; allow for follow-up and nonresponse in main plan
Returned questionnaires / consents10 / 10All consenting participants returned a form; very small sample
Usable forms / returned forms8 / 10 = 80%Criterion missed; inspect confusing required-item routing
Median completion time17 minutesCriterion missed; reduce burden and retest timing
Ambiguous frequency item6 of 10 respondents asked for clarificationReplace “regularly” with a defined recall period

Decision. Revise the item and form branching, remove an unnecessary question, then run a bounded retest of the revised survey. Do not interpret pilot correlations as confirmatory dissertation findings.

Complete dissertation-style pilot write-up — quantitative

A pilot of the planned online questionnaire was conducted with 40 eligible second-year undergraduate students who received an invitation through an authorised university channel. Ten students consented (25% of valid invitations), and all ten submitted questionnaires. Eight submissions met the predefined usability criterion (80%), below the planned 90% threshold. The median completion time was 17 minutes, exceeding the proposed 15-minute maximum. Six participants requested clarification of an item that used the term “regularly,” and review of the exports identified an unnecessary branching step. The item was revised to specify a defined recall period, and the redundant step was removed. Recruitment was considered workable for the setting, but survey usability and duration required a focused retest before full data collection. These pilot results informed the procedure; they were not used to test the substantive feedback-engagement hypothesis.

Why it works: the paragraph identifies the rehearsal, reports actual process denominators, compares them with criteria, documents revisions and clearly separates feasibility from substantive claims. In a real study, cite the approved protocol and report the actual numbers rather than adopting these invented teaching values.

Worked example 2: Qualitative interview pilot

Proposed main study (hypothetical). A researcher wants to explore how first-generation undergraduates describe seeking academic help. The pilot tests whether interview prompts are understandable and non-leading, whether consent and recording are workable, and whether participants can provide rich examples.

Pilot procedure. With required approval, three eligible volunteers are interviewed for 20–30 minutes. The interviewer records each prompt that produces only a yes/no answer, confusing terminology, interruptions and audio problems. Volunteers are invited to comment on difficult wording without being asked to endorse a preferred response.

Illustrative qualitative interview pilot problems, evidence and refinements
Observed problemIllustrative evidencePlanned refinement
Leading prompt“Did lecturers make you feel excluded?” elicited brief agreementAsk “Can you describe an experience of seeking help from a lecturer?”
Vague languageTwo participants asked what “institutional support” meantUse plain wording and context-specific follow-up probes
Thin accountsFirst responses often one sentenceAdd neutral “What happened next?” and “Can you give an example?”
Audio faultOne file contained significant background noiseMove to a quieter approved room; check equipment before consented recording

Complete dissertation-style pilot write-up — qualitative

A preliminary interview rehearsal was undertaken with three eligible volunteers after the necessary institutional permissions and consent processes. The exercise assessed prompt clarity, interview flow, recording quality and the ability of questions to invite concrete descriptions of help-seeking experiences. Two participants sought clarification of the term “institutional support,” and an early leading question produced short answers that did not explain the experience in detail. One audio recording contained substantial background noise. The guide was revised to use plain-language, open-ended prompts and neutral follow-up questions, while the recording procedure was changed to include an equipment check in a quieter approved setting. These changes were incorporated into the proposed main-study procedure, subject to any required approval amendment. The pilot was used to refine interview quality, not to establish themes representative of the broader population.

Why it works: it names the procedural evidence and revisions without treating three rehearsal interviews as sufficient evidence of the study’s substantive themes. It also respects consent and recording restrictions.

For more detail on turning a research question into neutral prompts, refer to Acadelyra’s qualitative interview guide.

Worked example 3: Clinic appointment-record feasibility pilot

Proposed main study (hypothetical). A public-health student intends to examine patterns in missed outpatient appointments using authorised retrospective clinic records. The pilot concerns record accessibility, identifier handling, whether the needed dates and appointment statuses are recorded, and whether the extraction template works. It does not contact patients or test an intervention.

Pilot procedure. After obtaining appropriate site and ethics/data-governance permissions, an authorised staff member extracts 25 purposively selected appointment records spanning both attended and missed appointments. The researcher works only with the approved minimum necessary fields, using a controlled data-transfer process. Twenty-five records are enough for this illustrative workflow check but not to estimate the clinic’s overall non-attendance rate precisely.

Illustrative results and decisions from a clinic appointment-record feasibility pilot
Feasibility checkIllustrative resultDecision
Appointment status present25 of 25 recordsField available for proposed definition
Appointment date present24 of 25 recordsClarify how a missing date will be handled
Duplicate encounter identifiers2 potential duplicatesVerify source-system extraction logic
Unapproved identifiers1 export initially included a direct identifierStop transfer, correct export settings and follow approved incident process as applicable
Secure transferApproved secure location availableContinue only after the corrected workflow is verified

Complete dissertation-style pilot write-up — records study

A feasibility review of the proposed appointment-record extraction was conducted under the applicable institutional and data-access approvals. An authorised custodian selected 25 appointment records representing attended and missed visits to test the extraction form and secure transfer. Appointment status was present for all 25 records, while one record lacked an appointment date. Two potential duplicate encounter identifiers required source-system clarification. A test export also contained a direct identifier outside the planned dataset; the transfer was halted and the export procedure corrected in line with approved data-governance requirements. Following these findings, the extraction template was updated to validate date completeness and duplicate identifiers before release to the analysis environment. The pilot supported refinement of the record workflow; the 25 selected records were not used to estimate the population prevalence of missed appointments.

Why it works: it treats data protection as part of feasibility, explains what the pilot uncovered, and does not generalise from a purposive set of records. The study must not begin until the required authorisations and corrected safeguards are in place.

How to write the pilot study section in your dissertation

A proposal describes what will be piloted and the criteria for deciding whether the procedure is ready. A completed dissertation should report what was actually done and what changed. Depending on departmental conventions, this material may sit in the methodology chapter, a feasibility subsection, an instrument-development section or an appendix. Keep the account transparent enough for readers to trace the decisions without reproducing every item of raw pilot feedback.

Components to report in a dissertation pilot-study section
ComponentQuestion to answerUseful detail
PurposeWhy was the pilot necessary?Specific recruitment, instrument or workflow uncertainty
Participants/dataWho or what was rehearsed?Eligibility, pilot count, setting and approval
ProcedureWhat steps were trialled?Invitation, consent, administration, extraction, timing
CriteriaHow was feasibility judged?Prespecified indicator, denominator and threshold
ResultsWhat worked and failed?Counts, durations, faults, comments, missingness
RevisionsWhat changed?Exact wording/process corrections and approval implications
DecisionWas a retest needed?Proceed, revise, retest or reconsider main-study feasibility

Weak versus strong pilot reporting

Weak. “A pilot study of ten students was conducted. The questionnaire was found reliable and valid, so the main research proceeded.”

This claims measurement quality without saying how it was checked, what the responses showed or whether ten responses could support those assertions. It hides the decisions readers need to evaluate.

Stronger. “The pilot tested item clarity, survey routing and completion time with ten consenting students. Six requested clarification of a frequency item, and median completion time exceeded the planned limit. The item was revised to use a defined timeframe, routing was simplified and a focused retest was planned. The pilot was not treated as a definitive test of scale validity or study hypotheses.”

The stronger version says what the pilot could establish, reports concrete observations and describes what followed. If measurement properties are being evaluated, explain the exact approach and its limits rather than claiming a scale is validated by a small convenience sample.

Can pilot data be included in the main study?

Sometimes, but not automatically. Eligibility, instrument revisions, consent, the study protocol and methodological comparability matter. If a pilot questionnaire was substantially changed, pooling responses with a later version may not be justified. If the pilot was a separate rehearsal with different consent terms, those records may not be authorised for substantive analysis. State the rule in advance where possible and obtain approval for any material change. Never silently merge pilot and main datasets.

Common pilot-study mistakes

  • Calling an informal practice run a completed pilot without defining its aim or recording observations.

  • Collecting participant data without the approvals, permissions or consent required for the activity.

  • Choosing a pilot sample using a universal number rather than the feasibility question.

  • Reporting only that “everything went well” without denominators, criteria or evidence.

  • Using an underpowered pilot p-value as proof the main hypothesis will succeed.

  • Changing questions after the pilot but failing to document what changed and why.

  • Claiming reliability or validity from a few participants without an appropriate measurement study.

  • Ignoring recruitment refusals, technical failures or missing observations.

  • Using pilot records in main analysis without addressing protocol and consent differences.

  • Failing to retest a critical workflow after making substantial corrections.

Pilot-study checklist

☐ The pilot has a specific feasibility purpose, not merely a label.

☐ The tested steps and the limits of the rehearsal are clear.

☐ All required ethics, site and data permissions are addressed.

☐ The pilot sample or records are justified for the feasibility question.

☐ Recruitment, consent and data handling are appropriately documented.

☐ Useful indicators and decisions were identified before results were reviewed.

☐ Results report accurate numerators, denominators and durations.

☐ Failures and inconvenient observations are reported honestly.

☐ Every important revision has a documented reason.

☐ Any revised procedure needing approval or retesting is identified.

☐ The distinction between feasibility and substantive hypothesis testing is maintained.

☐ The pilot and main-study datasets are not silently combined.

Frequently asked questions

What is the difference between a pilot and a feasibility study?

Terminology varies. A feasibility study focuses on whether a proposed investigation or intervention can be carried out; a pilot is often a smaller rehearsal of part or all of a future study. Explain your chosen meaning rather than assuming a universal distinction.

How many participants should a dissertation pilot include?

There is no single correct number. Justify it with the specific process being tested, the precision you need, recruitment constraints and departmental or ethics guidance. Ten interview rehearsals and ten estimates of a low response rate provide different kinds of evidence.

Can I conduct a pilot before ethics approval?

Do not assume so. Internal instrument review without human participants may be treated differently from recruiting volunteers or extracting identifiable records. Ask the appropriate ethics or governance office and follow its decision before collecting research data.

Should I calculate Cronbach’s alpha in a pilot?

You may examine internal consistency if this fits your measurement plan, but a small pilot estimate can be unstable and does not by itself establish validity. Clarify whether the pilot is for comprehension, feasibility, or a serious psychometric evaluation. Report estimates and their limitations honestly.

Do I have to include the pilot questionnaire or interview guide?

Many departments accept an appendix containing the instrument, selected changes or an audit trail. Include what your programme and ethics arrangements permit, and avoid attaching material that could expose participant identity or restricted content.

What if my pilot fails?

A pilot that uncovers a genuine problem can be very useful. Document the failure, identify its cause, decide what needs revision and whether a retest is necessary. A failed feasibility criterion is a reason to improve or reconsider the plan, not to disguise the result.

Final takeaway

A credible dissertation pilot study answers a practical question: can the proposed research process work well enough to justify proceeding? Define that question before testing; gather the relevant process evidence; compare it with justified criteria; document the changes; and state the remaining uncertainties. A pilot helps improve methods. It is not a shortcut to proving your dissertation hypothesis.

If your project still needs a coherent sequence of recruitment, instruments and secure data handling, review our data collection procedure guide before finalising the main study.

Need help understanding a research brief, planning methods or reviewing your methodology section? Acadelyra offers academic guidance and writing support. Keep your research decisions and submissions consistent with your institution’s academic-integrity rules.

You can request academic help and describe the specific part of your study you are working on.

Academic and methodological references

NIH NCCIH. Pilot Studies: Common Uses and Misuses.

NIH NCCIH. Clinical Trials Funding Opportunities FAQs (feasibility measures and sample justification).

Guidelines for Designing and Evaluating Feasibility Pilot Studies. Medical Care (2022).

What is a pilot or feasibility study? A review of current practice and editorial policy. (2010).