01 · The Question
How Small Can a Thesis Research Question Become and Still Be Defensible?
Suppose your preferred thesis contains several populations, outcomes, comparisons, methods, or studies. If time, recruitment, access, or resources deteriorate, what is the smallest version of the project that would still count as a worthwhile thesis?
This is a useful question even when nothing has gone wrong.
Identifying a minimum viable research question forces you to distinguish what the thesis genuinely needs from what merely makes it larger. The difficulty is that “minimum” can sound like lowering standards. That is not the objective. A defensible minimum still has to address a meaningful uncertainty, support appropriate analysis, produce claims justified by evidence, and satisfy the requirements of the degree.
03 · What You Need to Know
A Defensible Minimum Has Both a Floor and a Ceiling
“Minimum viable” is a planning concept, not an academic standard
Universities do not ordinarily define thesis quality through a concept called a minimum viable research question. The phrase is useful here as a planning device: identify the smallest defensible version of the inquiry before optional complexity is added.
The actual academic threshold comes from your degree regulations, disciplinary standards, examination criteria, and the methodological requirements of the research.
That distinction matters. You should not tell an examiner that the thesis is “minimum viable” and expect applause for project management. The concept is for designing a robust project, not redefining what your degree requires.
The question must still be worth answering
Feasibility alone establishes only that a study can be done. It does not establish that it should be done.
Frameworks such as FINER make this distinction explicit by evaluating research questions for feasibility alongside interest, novelty, ethics, and relevance. Contemporary guidance similarly emphasizes that good research questions should produce results that are both achievable and useful.
A minimum viable question therefore needs a substantive core. It might resolve an uncertainty, test an assumption, explain a process, examine a meaningful relationship, interpret evidence, evaluate a theoretically motivated proposition, or otherwise add something defensible to an existing scholarly conversation.
If narrowing leaves only an obvious descriptive fact with little analytical consequence, you may have crossed below the useful minimum.
The question must be answerable with evidence you can actually obtain
The Chinese University of Hong Kong's thesis guidance describes a research question as needing to be supportable, meaning that data, materials, or other evidence capable of answering it can be obtained within the scope of the study. It also describes a good question as specific and manageable.
That gives the minimum viable question an upper constraint: its evidentiary demands cannot exceed the researcher's realistic capacity.
Ask what evidence would constitute a credible answer. Then verify whether that evidence is accessible within the thesis timeline, resources, permissions, expertise, and methodological environment.
The minimum concerns the question, not an arbitrary amount of data
There is no universal minimum number of participants, interviews, institutions, variables, hypotheses, chapters, or research questions that makes a thesis defensible.
Sample adequacy depends on methodology. A statistical analysis may require enough observations to achieve appropriate precision or power under the intended design. A qualitative study may require sampling and information sufficient for its particular analytical purpose. Historical, theoretical, computational, mathematical, and design-based theses operate under entirely different evidentiary logics.
Therefore, “What is the minimum sample for a thesis?” cannot be answered independently of the research question and method.
A useful minimum preserves one consequential uncertainty
Imagine a broad project examining whether an educational intervention works, why it works, for whom it works, how participants experience it, whether effects persist, and whether implementation differs among institutions.
A minimum viable version would not simply perform all six tasks badly.
Instead, identify which uncertainty constitutes the thesis's central contribution. Perhaps the important question concerns whether the intervention changes one specified outcome under a defined set of conditions. The other questions may become extensions rather than requirements.
Minimum viable question
Removes optional breadth while preserving a meaningful uncertainty, appropriate evidence, and sufficient analytical depth.
Trivial question
Achieves feasibility by removing so much uncertainty or analytical substance that little worthwhile scholarly work remains.
The minimum should retain the logic of the research design
You cannot make a study viable merely by shrinking the sample, dropping inconvenient observations, or deleting a comparison after the original design becomes difficult.
If the question asks whether two groups differ, both groups are logically necessary. If it asks about change over time, temporal evidence is necessary. If the contribution depends on understanding a process, evidence that only measures an outcome may be insufficient.
The minimum viable version must remain methodologically capable of answering the revised question. When necessary evidence is removed, the question itself has to change.
Remove optional breadth in a deliberate order
When searching for the minimum, it helps to distinguish components according to the work they perform.
| Component |
Keep when... |
Candidate for removal when... |
| Additional population |
Comparison is central to the contribution |
It merely broadens coverage |
| Additional site |
Contextual variation is theoretically necessary |
It mainly increases generality or convenience |
| Additional outcome |
It answers an essential part of the question |
It is interesting but peripheral |
| Additional method |
It provides evidence the other method cannot |
It adds richness without changing the central inference |
| Additional time point |
Change or persistence is central |
The question does not depend on temporal development |
| Additional subquestion |
It is necessary to answer the central question |
It could become an independent future study |
This approach is closely related to reducing an ambitious thesis without making it trivial. The difference is that the minimum viable question deliberately asks where further reduction must stop.
One central question may be enough
A thesis does not become academically substantial because it contains many numbered research questions. One well-designed central question can generate substantial literature review, methodological reasoning, evidence, analysis, and discussion.
Subquestions are useful when they decompose the central problem into necessary components. They are less useful when they merely create additional topics to investigate.
When identifying the minimum viable question, ask whether every subquestion is necessary to answer the central one. If not, it may belong outside the minimum architecture.
The minimum can use existing data
If a suitable existing dataset, archive, corpus, repository, or administrative source can answer the central question, primary data collection is not automatically required to make the thesis legitimate.
Using existing evidence may substantially reduce recruitment and collection risk while preserving analytical depth. The important condition is that the data fit the question and support the intended inference.
Whether existing data provide a stronger thesis strategy than collecting new data therefore depends on evidentiary suitability rather than on assumptions about which approach involves more effort.
The minimum should survive foreseeable failure better than the preferred design
One reason to identify a minimum viable question is contingency planning.
Suppose your preferred project includes three institutions, but the central contribution can be established within one appropriate institution and the other two primarily test generality. Knowing this before recruitment begins gives you a defensible boundary if access changes.
By contrast, if all three institutions are essential to the comparison embedded in the research question, a one-institution fallback is not the same study. You may need a genuinely different question.
This is why completion risk should be considered when the research question is chosen, not merely when the project begins to struggle.
A minimum viable question can protect ambition
This may sound paradoxical, but establishing the minimum can make responsible expansion easier.
Once you know what must be completed for the thesis to remain defensible, additional components can be evaluated as deliberate extensions. You might add a second population to examine a boundary condition, a follow-up measurement to investigate persistence, or interviews to explain an unexpected quantitative pattern.
The difference is that optional ambition is now distinguishable from the project's load-bearing structure.
The degree determines the floor
A question suitable for an undergraduate thesis may not satisfy a master's program. A master's question may not support the original contribution expected of a doctorate. Expectations also vary substantially among disciplines and institutions.
Therefore, the minimum viable question must be calibrated to the actual degree. Generic advice can identify feasibility and research-design problems, but only your program's current regulations, disciplinary expectations, and examination standards can determine whether the proposed contribution is sufficient.
04 · A Practical Example
Finding the Minimum Defensible Version of a Thesis
Hypothetical Example
A master's thesis on AI-supported feedback
A student initially proposes to compare generative AI feedback with instructor feedback across three universities, measure writing performance and motivation, conduct student interviews, and repeat the assessment after three months.
The preferred project could provide useful evidence, but several components create substantial recruitment, access, coordination, and analytical demands.
State the central uncertainty The student primarily wants to know whether a specified form of AI-supported feedback during revision is associated with subsequent independent writing performance.
Identify essential evidence The study needs a defined AI-supported revision condition, an appropriate measure of subsequent independent performance, and a design capable of supporting the intended comparison or inference.
Remove optional breadth Institutional comparison, motivation, interviews, and three-month follow-up are removed from the minimum because none is necessary to answer the central question as revised.
Check the remaining contribution The question still addresses a meaningful distinction between performance achieved with AI assistance and subsequent performance without that assistance.
Define the minimum The minimum viable project investigates that relationship within one appropriate and accessible student population using defensible measures and analysis.
The student may still choose to add interviews or follow-up measurements if resources permit and if they strengthen the contribution. The important point is that the thesis no longer requires those additions merely to remain academically coherent.
07 · A Quick Checklist
Find the Minimum Defensible Research Question
Before defining your minimum viable thesis, check:
State the consequential uncertainty that must remain after the project is narrowed.
Explain why answering that question would still be worthwhile within the relevant scholarly literature.
Identify the minimum evidence methodologically required to support a credible answer.
Remove populations, sites, outcomes, variables, methods, and subquestions that do not contribute necessary evidence.
Verify that the remaining data, participants, materials, expertise, and resources are realistically accessible.
Confirm that the narrowed question still requires meaningful analysis, interpretation, or scholarly reasoning.
Align the intended claims with the narrower population, context, evidence, and design.
Check the proposed minimum against current degree regulations, disciplinary expectations, and supervisory judgment.
Label any additional studies, populations, outcomes, or methods as extensions and state exactly what contribution each would add.