03 · What You Need to Know
Determining Whether Independent Feasibility Research Is Justified
What Is a Separate Feasibility Study?
A separate feasibility study is an investigation conducted before the intended definitive study to examine whether and how that future research can be undertaken.
Its primary purpose is to inform decisions about the proposed study rather than answer the definitive research question.
For example, researchers planning a large educational intervention might conduct preliminary research to determine whether participating institutions can recruit sufficient classes, deliver the intervention, and collect the required outcomes.
The feasibility investigation has its own objectives, procedures, findings, and conclusions. Its results inform the subsequent research design.
In the randomized trial literature, feasibility studies may be conducted without rehearsing the entire proposed trial. Pilot studies, by contrast, are commonly understood as a subset of feasibility research that implements the future study or selected components on a smaller scale.
This distinction is developed in the methodological framework of Eldridge and colleagues.
What Makes a Feasibility Study Separate?
A separate feasibility study is conducted as a distinct preliminary investigation rather than as an initial phase incorporated into the definitive study.
It may use different participants, procedures, outcomes, or data collection arrangements from the future study.
For example, researchers may interview teachers and administrators about the practicality of implementing a proposed instructional intervention before designing the definitive evaluation.
These interviews may generate useful feasibility evidence without becoming part of the eventual effectiveness dataset.
Separate feasibility research should not be confused automatically with an external pilot. An external pilot specifically rehearses the proposed definitive study or its components, whereas a separate feasibility study may investigate narrower questions without doing so.
When Is a Separate Feasibility Study Most Strongly Justified?
The strongest justification arises when consequential uncertainty remains and proceeding directly to the definitive study would expose the project to avoidable methodological, operational, financial, or ethical risks.
For example, a proposed trial may depend on recruiting schools that have not yet agreed to participate. If access is uncertain and alternative sites are limited, feasibility research could substantially influence whether the trial is possible.
Similarly, a complex intervention may require resources or staff capabilities that have not been demonstrated in the intended setting.
A separate study may be warranted when the uncertainty is substantial enough that the main research design cannot responsibly be finalized without further evidence.
However, the mere existence of uncertainty does not establish necessity. Research inevitably involves uncertainty, and not every question requires an independent preliminary investigation.
How Do You Distinguish Critical Uncertainty From Ordinary Research Risk?
Ordinary research risks can often be managed through established procedures, contingency planning, or existing methodological evidence.
Critical uncertainty concerns a condition whose failure could make the study infeasible, invalidate essential comparisons, or prevent collection of the evidence required to answer the research question.
| Situation |
Potential response |
Reason |
| An established questionnaire needs a technical check. |
Routine verification or focused pretesting |
The uncertainty may be narrow and readily addressed. |
| Access to the target population has not been established. |
Focused feasibility investigation |
Recruitment may be impossible without institutional cooperation. |
| A new intervention requires several unfamiliar resources. |
Separate feasibility study or pilot |
Implementation uncertainty may affect the entire definitive design. |
| Longitudinal retention is unknown in a difficult-to-reach population. |
Feasibility assessment with appropriate follow-up |
Attrition may undermine the planned outcome analysis. |
| Recruitment and procedures are well established in comparable settings. |
Proceed with justified safeguards |
Existing evidence may be sufficient. |
| The proposed study requires an untested participant-level data linkage. |
Technical verification and, if necessary, separate feasibility research |
Failure could make the intended analysis impossible. |
The appropriate response depends on the quality of existing evidence, not simply the apparent complexity of the procedure.
When Does Recruitment Uncertainty Justify a Separate Study?
Recruitment feasibility may warrant independent investigation when the required population is difficult to identify, access, screen, or enroll.
For example, researchers may intend to recruit university teachers with experience implementing a particular AI-based assessment system. The number of eligible teachers may be unknown, and institutional access may depend on several administrative approvals.
A feasibility study could estimate the available recruitment pool, examine willingness to participate, and identify barriers to access.
However, recruitment difficulties do not automatically require a separate study. A limited recruitment pilot or analysis of existing institutional records may answer the necessary questions.
The decision should reflect whether the information obtained could change the definitive sample, recruitment strategy, timeline, or decision to proceed.
See How Do You Decide Whether Recruitment Is Feasible Before Committing to the Main Study?.
When Does Intervention Complexity Justify Separate Feasibility Research?
Complex interventions may depend on several interacting components, such as staff training, participant engagement, technical systems, institutional policies, and implementation schedules.
For example, an AI-supported learning intervention may require teachers to review generated feedback, maintain instructional oversight, and coordinate feedback delivery with existing course requirements.
If these arrangements have not been demonstrated in the intended setting, feasibility research may help identify whether implementation is realistic.
The investigation might examine staffing, technical infrastructure, workload, and stakeholder acceptability before the definitive evaluation is finalized.
However, complexity alone is not sufficient justification. Existing implementation evidence may already address many relevant uncertainties.
When Does Participant Acceptability Require Independent Investigation?
Acceptability becomes particularly consequential when participation or intervention delivery depends on activities that individuals may regard as burdensome, intrusive, inappropriate, or difficult to accommodate.
For example, teachers may support the educational purpose of a proposed intervention while objecting to extensive classroom recording or additional reporting requirements.
A feasibility study may investigate these concerns through interviews, observations, or other suitable methods.
However, acceptability should not be reduced to whether participants express general approval.
Researchers may need to examine anticipated burden, perceived appropriateness, ethical concerns, and willingness to participate under realistic conditions.
Favorable attitudes do not guarantee subsequent enrollment or adherence.
When Do Measurement Problems Justify a Separate Feasibility Study?
Measurement uncertainty may warrant preliminary investigation when the proposed outcomes cannot yet be collected consistently or meaningfully.
For example, researchers may wish to evaluate changes in student engagement using learning analytics, questionnaires, and performance assessments.
They may need to determine whether relevant records can be accessed, whether the instruments are suitable for the population, and whether repeated assessments are practical.
However, these concerns should be distinguished.
Technical data access may require a focused systems test. Questionnaire comprehension may require cognitive interviewing. Establishing measurement validity may require a dedicated measurement study.
A broad feasibility study is justified when several consequential uncertainties must be investigated together, not merely because the researcher has not yet tested an instrument.
When Is Institutional or Organizational Feasibility the Main Concern?
Some studies depend on organizational arrangements that researchers cannot control directly.
Examples include institutional permissions, staff availability, access to administrative records, coordination across sites, and compliance with relevant governance requirements.
A university may agree in principle to participate in a study while lacking the operational capacity to implement the proposed procedures.
Feasibility research may clarify whether necessary commitments can be obtained and maintained.
However, consultation should not be mistaken for formal authorization. Ethical approval, data access permission, and other required agreements must still be obtained through the applicable processes.
Should Resource Uncertainty Trigger a Separate Study?
Potentially. Resource feasibility includes staffing, funding, equipment, technical support, time, and other requirements necessary to conduct the proposed research.
For example, a multisite intervention may require more teacher training and technical assistance than the research team originally anticipated.
A separate feasibility study can estimate resource demands under realistic conditions and identify requirements that could make the definitive study impractical.
However, a resource assessment does not always require a formal empirical study. Existing cost information, institutional consultation, or operational modeling may be sufficient.
The appropriate approach depends on the uncertainty and the importance of the resulting decision.
When Is a Separate Study Preferable to a Pilot?
A separate feasibility study that does not rehearse the definitive design may be preferable when researchers first need to establish whether the proposed study is possible or acceptable.
For example, conducting a randomized pilot may be premature if researchers have not established whether enough institutions will permit recruitment.
Similarly, a full intervention rehearsal may be unnecessary when the central uncertainty concerns the availability of qualified staff.
By contrast, a pilot may be more appropriate when the research team needs to examine how recruitment, allocation, intervention delivery, and outcome assessment function together.
These approaches are related rather than mutually exclusive. A project may begin with focused feasibility research and subsequently conduct a pilot if substantial operational uncertainty remains.
For further comparison, see When Is a Pilot Study More Appropriate Than a Feasibility Study?.
Can Feasibility Questions Be Addressed Without a Separate Study?
Yes. Some feasibility questions can be answered through document review, consultation, technical testing, small-scale pretesting, or analysis of existing operational information.
For example, researchers may confirm that an institution can export required learning management system records using synthetic data and approved technical procedures.
That activity may establish technical feasibility without requiring participant recruitment.
However, the adequacy of such evidence depends on the question being answered.
A successful technical demonstration cannot establish that students will accept an intervention or complete repeated assessments.
Can an Internal Pilot Replace a Separate Feasibility Study?
In some trial designs, an internal pilot can evaluate feasibility during an initial phase of the definitive study.
This approach may be efficient when the principal procedures are sufficiently established and progression decisions can be made using prespecified criteria.
However, an internal pilot is less suitable when substantial changes to the intervention, outcomes, eligibility criteria, or allocation procedures are likely.
Such changes may affect whether initial-phase data can be incorporated into the final analysis.
The choice requires careful methodological planning and, where relevant, statistical expertise.
An internal pilot should not be used merely to avoid acknowledging unresolved design problems.
How Do Ethical Considerations Affect the Decision?
Researchers should consider whether proceeding directly to a definitive study would expose participants to avoidable burdens or risks when the study's basic feasibility remains uncertain.
For example, a complex intervention may require considerable participant time even though the research team has not established whether essential outcome data can be collected.
Preliminary feasibility work may help prevent unnecessary participation in an investigation that cannot answer its intended question.
However, feasibility studies involving human participants also require appropriate ethical review and safeguards.
The preliminary designation does not exempt researchers from applicable requirements.
Does a Large or Expensive Study Automatically Require Separate Feasibility Research?
No. Large studies may benefit substantially from feasibility evidence, but their size alone does not establish that a separate study is necessary.
A large investigation using established procedures and reliable recruitment pathways may have fewer unresolved feasibility questions than a small but highly complex intervention.
Researchers should examine the probability and consequences of procedural failure, the availability of relevant evidence, and the potential value of additional preliminary information.
The justification should focus on consequential uncertainty rather than project size or perceived prestige.
What Should a Separate Feasibility Study Actually Measure?
The outcomes should correspond to the uncertainties that motivate the investigation.
Possible outcomes include recruitment yield, enrollment speed, participant retention, intervention adherence, data completeness, acceptability, implementation fidelity, and resource use.
However, not every study requires all these measures.
For example, a feasibility investigation focused on institutional access may prioritize permission processes and recruitment capacity rather than intervention adherence.
Outcomes should be defined clearly enough to support decisions about the definitive study.
The broader process of selecting feasibility questions is discussed in What Feasibility Questions Should You Answer Before Conducting a Definitive Study?.
Should a Separate Feasibility Study Have Progression Criteria?
When feasibility findings will determine whether the definitive study proceeds, prespecified progression criteria may improve transparency and decision-making.
Criteria can identify conditions supporting progression, modification, or reconsideration.
For example, a recruitment criterion might reflect the minimum enrollment rate required to complete the definitive study within available resources.
However, thresholds should be justified by the proposed research rather than adopted as universal standards.
Progression decisions should also consider uncertainty, contextual information, and whether identified problems can be remedied.
Can a Feasibility Study Be Too Small?
Yes, if its scale is insufficient to answer the feasibility questions it was designed to investigate.
For example, interviewing a few stakeholders may reveal important implementation concerns but provide limited evidence about the proportion of institutions likely to participate.
Similarly, a very small recruitment assessment may produce an imprecise enrollment estimate.
The sample size should reflect the purpose of the feasibility investigation, the required precision, and the relevant variation across participants or settings.
There is no universal sample size appropriate to every feasibility study.
What If the Feasibility Study Identifies Major Problems?
The research team should determine whether the problems can be addressed without undermining the definitive research question.
For example, inadequate recruitment may be improved by adding sites, extending the recruitment period, or revising communication procedures.
However, some changes may alter the target population or intervention sufficiently to require substantial redesign.
A feasibility study may support proceeding, proceeding with modifications, conducting further preliminary work, or deciding not to proceed.
The appropriate decision depends on the evidence and the consequences of unresolved limitations.
Does a Successful Feasibility Study Guarantee That the Main Study Will Work?
No. Feasibility findings are conditional on the participants, settings, procedures, and resources examined.
A study that appears feasible in one university may encounter different conditions when expanded to several institutions.
Recruitment rates may change, staff availability may decline, or technical systems may behave differently at scale.
Researchers should interpret feasibility evidence as informing the probability and conditions of successful implementation, not as eliminating uncertainty.
How Should the Need for Separate Feasibility Research Be Justified?
A proposal should identify the specific uncertainties, explain why existing evidence is insufficient, and show how the feasibility findings will influence the definitive study.
For example, a researcher might explain that the proposed intervention requires institutional coordination and technical infrastructure whose availability has not been established.
The feasibility study would then investigate those requirements before finalizing the definitive evaluation.
A strong justification connects the preliminary objectives to consequential decisions.
Simply stating that feasibility research is necessary because pilot studies are good practice provides limited methodological justification.
Watch Out
Do not conduct a separate feasibility study merely as a procedural formality, and do not proceed directly to the main study simply because preliminary work would delay publication. The decision should reflect whether unresolved uncertainty could materially compromise the research and whether additional evidence would improve that decision.
