03 · What You Need to Know
Determining Whether Revised Research Procedures Need Further Feasibility Testing
Why Might Research Redesign Require Another Feasibility Study?
Feasibility evidence is generated under particular conditions, including the study population, research procedures, setting, resources, and implementation arrangements.
When these conditions change, the original findings may no longer provide sufficient evidence about the revised study.
For example, a feasibility study may demonstrate that teachers can deliver a four-week intervention. If the research team subsequently extends the intervention to an entire semester, the original findings may provide limited information about sustained workload and participant adherence.
Similarly, recruitment feasibility established at one university may not fully apply when the definitive study moves to institutions with different populations or administrative procedures.
Further testing may therefore be necessary when the redesign changes the conditions underlying the original feasibility conclusions.
Does Every Research Modification Require Repeating the Study?
No. The need for additional testing depends on the significance of the change and its consequences for the definitive research.
For example, correcting a typographical error in participant instructions may require careful review but not a new feasibility study.
Changing the primary outcome instrument, however, may introduce uncertainty about administration, scoring, and the suitability of the resulting measurements.
Researchers should distinguish minor corrections from changes that affect essential procedures or the study's inferential purpose.
The relevant question is not simply how many modifications were made, but whether they invalidate or weaken the evidence supporting progression.
What Is the Difference Between Verification, Targeted Retesting, and Repeating a Feasibility Study?
These approaches address different levels of uncertainty.
| Approach |
Purpose |
When it may be appropriate |
| Procedural verification |
Confirm that a specified correction functions as intended |
Minor technical or administrative changes |
| Targeted feasibility retesting |
Evaluate one or more consequential modified procedures |
Focused changes to recruitment, measurement, or delivery |
| Repeated feasibility study |
Reassess several connected feasibility requirements under revised conditions |
Substantial redesign affecting multiple essential procedures |
| New pilot study |
Rehearse the revised definitive study or selected components on a smaller scale |
Major changes to how connected research procedures operate |
These categories describe practical options rather than universally standardized study classifications.
The appropriate approach should be justified by the uncertainties remaining after redesign.
When Is Targeted Retesting Sufficient?
Targeted retesting may be sufficient when the original feasibility study established that most procedures work and the redesign affects a limited component.
For example, a study may demonstrate adequate recruitment and intervention delivery but identify an error in the data export process.
After correcting the export procedure, researchers may verify that values, identifiers, and missing-data codes are preserved accurately.
Repeating recruitment and intervention delivery would contribute little if those procedures remain unchanged and the original evidence remains relevant.
However, targeted testing should examine relevant interactions with other procedures when a correction could affect them.
When Is a Complete Repeat More Appropriate?
A broader feasibility investigation may be justified when the redesign substantially changes several interconnected procedures.
For example, replacing a classroom intervention with an asynchronous online intervention may alter recruitment, delivery, adherence, assessment, and technical support requirements.
In that situation, the original feasibility findings may no longer address the conditions under which the revised study will operate.
Researchers may need to reassess multiple domains rather than verify one isolated change.
However, even substantial redesign does not necessarily invalidate every previous finding.
The new investigation should focus on the uncertainties that remain rather than repeat activities without a clear purpose.
How Do You Determine Whether Existing Feasibility Evidence Still Applies?
Researchers should compare the original and revised designs systematically.
Important dimensions include the target population, intervention, setting, recruitment procedures, measurement arrangements, analysis requirements, and available resources.
For each dimension, ask whether the original feasibility finding depended on a condition that has changed.
For example, evidence that students can complete an online questionnaire may remain relevant if only the recruitment message changes.
However, evidence about questionnaire completion may be less applicable if the assessment becomes substantially longer or is moved to a more demanding point in the academic calendar.
The goal is to determine which conclusions remain supported and which require new evidence.
What Changes to Recruitment May Require Retesting?
Recruitment modifications may require additional testing when they change the population reached, eligibility criteria, recruitment channels, or institutional arrangements.
For example, adding universities may increase the accessible population but introduce different administrative requirements.
Changing eligibility criteria may also affect the characteristics of participants and the interpretation of the definitive findings.
Researchers should distinguish expanding access to the same target population from redefining the target population itself.
Where recruitment capacity remains uncertain, targeted assessment may be appropriate.
The original recruitment feasibility evidence should not be extrapolated automatically to substantially different recruitment conditions.
Should You Retest After Changing the Intervention?
Potentially, especially when modifications affect essential intervention components, delivery requirements, or participant experience.
For example, reducing the frequency of feedback activities may make an intervention easier to deliver but also change its intended educational mechanism.
Researchers should identify whether the modification preserves the intervention's core function.
If the revised intervention differs substantially from what was previously tested, additional feasibility evidence may be needed.
However, minor adjustments to scheduling or instructions may require less extensive verification when the essential intervention remains unchanged.
What If the Redesign Changes Intervention Dose?
Changes to intervention duration, frequency, or intensity may affect adherence, workload, acceptability, and implementation fidelity.
For example, a six-week intervention may be shortened to four weeks after participants report excessive burden.
The revised schedule may be more manageable, but the original feasibility evidence does not necessarily establish that participants will engage with the new arrangement as intended.
Researchers should also consider whether the changed dose remains theoretically appropriate.
Feasibility testing can examine implementation, but it cannot automatically establish that the revised dose produces the intended outcome.
When Does Changing the Outcome Measure Require Retesting?
Replacing or substantially modifying an outcome measure may introduce new uncertainties about administration, participant understanding, scoring, and data completeness.
For example, a study may replace a lengthy engagement questionnaire with a shorter instrument to reduce participant burden.
The revised instrument may be easier to administer, but its suitability for the intended construct interpretation must still be considered.
A feasibility assessment may examine completion and administration, while measurement validity may require separate evidence.
Researchers should not assume that a shorter or more convenient instrument is automatically equivalent to the original.
Should Changes to Data Collection Technology Be Retested?
Yes, when the changes affect consequential data capture, transfer, storage, or analysis procedures.
For example, replacing one survey platform with another may alter skip logic, variable coding, date formats, or export structures.
Technical testing using synthetic records can verify many of these functions.
However, real-world usability or accessibility concerns may require additional testing with appropriate participants.
The scope should reflect the potential consequences of technical failure.
Researchers should not repeat the entire feasibility study merely because a software interface changes, unless the change affects broader research procedures.
What If the Redesign Changes the Target Population?
Changing the target population can substantially affect the applicability of previous feasibility evidence.
For example, a study initially developed for postgraduate students may be redesigned for first-year undergraduates.
Recruitment, participant understanding, intervention burden, and outcome assessment may differ between these groups.
Researchers should determine whether the original evidence remains relevant.
Where the revised population differs in consequential ways, additional feasibility assessment may be justified.
The research question and intended generalizations may also need revision.
What If the Study Moves to Another Institution?
Institutional changes may affect access, staffing, infrastructure, scheduling, governance, and participant characteristics.
For example, an intervention that functions well at a university with extensive technical support may be difficult to implement at an institution with more limited infrastructure.
Researchers should examine which feasibility findings depend on the original institutional context.
However, moving to another institution does not automatically require repeating every preliminary procedure.
Existing evidence may remain applicable to components that are sufficiently similar.
Targeted testing may address the specific differences.
Does Changing the Research Design Require Another Pilot?
Substantial changes to the research design may justify a new pilot, particularly when they alter how participants are recruited, assigned, assessed, or analyzed.
For example, changing from individual randomization to cluster randomization introduces new questions about cluster recruitment, allocation, contamination, and statistical structure.
The original pilot may not have tested these procedures.
A new pilot may therefore be appropriate to rehearse the revised design.
However, researchers should identify the specific uncertainties rather than assume that every design modification requires a full repetition.
How Does the Decision Differ for Internal and External Pilots?
An external pilot is conducted separately from the definitive study. Its findings may support substantial redesign before the definitive research begins.
An internal pilot forms an initial phase of the definitive study, with prespecified progression arrangements.
For an internal pilot, major modifications may affect whether the initial observations remain compatible with the final analysis.
Researchers should consider the protocol, statistical implications, and relevant oversight requirements before changing essential procedures.
A substantial redesign may require treating the revised investigation as a different study rather than simply continuing the original internal pilot.
Should You Repeat Feasibility Testing After an Amber Progression Decision?
Not automatically. Amber commonly indicates that identified problems may be addressed through specified modifications.
Some corrections can be verified directly, while others require additional empirical evidence.
For example, correcting a known database configuration error may be confirmed through technical tests.
However, introducing a new recruitment strategy may require observation of actual enrollment to establish whether it achieves the intended rate.
The meaning of an amber decision depends on the progression framework and corrective actions established for the study.
Should You Repeat Testing After a Red Progression Decision?
A red classification may indicate that the current design cannot proceed without substantial changes.
If the research team develops a credible alternative design, further feasibility research may be appropriate.
However, repeating the original study without addressing the causes of infeasibility is unlikely to provide a useful solution.
Researchers should identify what has changed and why the revised procedures are expected to overcome the previous barriers.
Where no realistic corrective pathway exists, a decision not to proceed may be more defensible than repeated preliminary testing.
How Should Progression Criteria Be Updated After Redesign?
Progression criteria should be reconsidered when the definitive study's requirements change.
For example, adding recruitment sites may alter the expected recruitment rate per institution while preserving the overall enrollment target.
Changing the primary outcome may require a different completeness criterion.
However, researchers should not revise thresholds solely to classify previous unfavorable findings as acceptable.
The revised criteria should be justified by the new design and documented before interpreting new feasibility evidence where possible.
The relationship between progression criteria and definitive-study requirements should remain explicit.
Can Previous Feasibility Data Be Reused?
Yes, when the data remain relevant to the revised feasibility questions and their use is methodologically and ethically appropriate.
For example, previous observations about teacher training time may inform planning if the training procedures remain unchanged.
However, estimates derived under substantially different conditions should not automatically be pooled with new observations.
Researchers should distinguish evidence that remains applicable from evidence that no longer represents the revised study.
Any reuse should be consistent with consent, data governance, and applicable protocol requirements.
Can Old and New Feasibility Results Be Combined?
Sometimes, but combining results requires attention to comparability.
For example, recruitment rates from two phases may be summarized together when the populations, procedures, and observation periods are sufficiently comparable.
However, pooling rates from substantially different recruitment strategies may conceal important variation.
Similarly, intervention adherence before and after major redesign may represent different activities.
Researchers should document the changes and consider reporting results separately when combining them would produce misleading summaries.
How Much Additional Testing Is Enough?
The scope should reflect the uncertainty that remains and the evidence needed for a progression decision.
A technical correction may require a defined set of verification tests.
A revised recruitment strategy may require an observation period sufficient to assess enrollment capacity.
A substantially modified intervention may require a broader pilot examining delivery, adherence, acceptability, and outcome collection.
There is no universal sample size or duration for repeated feasibility testing.
Researchers should justify the scope according to the revised study's requirements rather than repeat the original sample size automatically.
Should the Same Participants Be Used in Repeated Testing?
It depends on the research question, the procedures being tested, and the relevant ethical arrangements.
Returning participants may provide useful information about how revised procedures differ from the original experience.
However, familiarity with the intervention may influence their responses or behavior.
For example, students who have already completed an instructional activity may find a revised version easier than new participants would.
Where the objective is to evaluate first-time usability or recruitment, a new sample may be more informative.
Researchers should justify participant selection and obtain appropriate consent or approvals.
Can Repeated Feasibility Studies Become Methodologically Problematic?
Yes, particularly when researchers repeatedly modify procedures until favorable results appear without a transparent decision framework.
Repeated testing may also consume resources while delaying the definitive investigation.
For example, a team may continue changing recruitment criteria after every disappointing enrollment estimate, eventually studying a population substantially different from the original target.
Such changes may undermine the scientific purpose of the research.
Researchers should establish what uncertainty each additional test is intended to resolve and what decision will follow from the results.
Repetition should not become an indefinite search for a favorable feasibility classification.
What Is the Risk of Overfitting the Study to the Pilot Setting?
Researchers may inadvertently redesign procedures to work exceptionally well in the particular setting or sample used during preliminary testing.
For example, an intervention may become highly dependent on a small group of enthusiastic teachers who receive unusually intensive support.
That arrangement may not be sustainable in the definitive study.
Researchers should consider whether modifications improve general implementation feasibility or merely accommodate unusual features of the pilot setting.
Where appropriate, testing across different sites or implementers may provide additional evidence about transferability.
Should Researchers Reassess Ethical Requirements After Redesign?
Yes, when modifications affect participant procedures, risks, consent arrangements, data handling, or other matters subject to ethical oversight.
For example, introducing additional digital monitoring may change the information collected and the privacy implications of participation.
Researchers should determine whether amendments or additional approvals are required under the applicable institutional procedures.
However, not every administrative correction necessarily requires a new ethics application.
The relevant requirements should be verified with the responsible review body.
Feasibility testing should not proceed under outdated approvals when consequential changes require review.
How Should Researchers Document the Redesign?
Documentation should identify the original feasibility problem, the proposed modification, its rationale, and the evidence supporting the revised approach.
Researchers should also specify which original feasibility findings remain applicable and which questions require further testing.
For example, a change log may distinguish revised recruitment procedures from unchanged outcome assessments.
Where new testing is conducted, its objectives and progression criteria should be established clearly.
Consequential changes should be reflected in the protocol and relevant approvals.
This documentation supports transparency and helps prevent retrospective reinterpretation of the research process.
How Should Repeated Feasibility Findings Be Reported?
The report should explain why additional testing was undertaken and what changed between the original and revised procedures.
Researchers should distinguish original findings from results obtained after modification.
Where outcomes are compared, differences in participants, settings, measurement, or implementation should be acknowledged.
The report should explain whether the new evidence resolves the uncertainty motivating redesign.
It should also identify limitations and remaining conditions for progression.
Successful retesting should not be described as proof that the definitive study will be free from implementation problems.
When Should Researchers Stop Repeating Feasibility Studies?
Further testing becomes difficult to justify when the relevant uncertainty has been resolved sufficiently for a progression decision or when additional testing is unlikely to change the decision.
For example, repeated recruitment assessments may be unnecessary when the study has established that the required eligible population is unavailable and no suitable alternative exists.
Conversely, additional testing may also be unnecessary when a corrected procedure has been verified adequately and no consequential uncertainty remains.
The decision should consider the expected value of further evidence relative to its costs, burdens, and likely influence on the definitive study.
Researchers should be prepared to proceed, redesign, or discontinue rather than remain indefinitely in preliminary research.
Watch Out
Do not repeat a feasibility study automatically after every modification, and do not assume that previous feasibility evidence remains valid after substantial redesign. The appropriate response is to identify what changed, determine which uncertainties remain, and test only what is necessary to support a defensible progression decision.
