03 · What You Need to Know
How to Find and Evaluate a Meaningful Research Problem
A Research Topic Is Not the Same as a Research Problem
A topic identifies an area you want to study. A research problem identifies something within that area that requires investigation.
Research topic
A subject or area of interest, such as generative AI in higher education.
Research problem
A specific uncertainty, difficulty, inconsistency, knowledge deficiency, or unresolved issue within that area that provides a reason for investigation.
“Student use of generative AI” is therefore a topic, not yet a research problem. You still need to establish what about that phenomenon is insufficiently understood and why gaining that understanding would matter.
The research problem eventually provides the rationale for your research question. The question expresses more precisely what you will try to find out, while the problem explains why there is something worth finding out in the first place.
Start With Something That Is Genuinely Uncertain
Research problems can emerge in different ways. Existing studies may produce conflicting findings. A theory may fail to explain an observation. A method may have an important limitation. Practitioners may repeatedly encounter a difficulty without adequate evidence about how to address it. A new technology, intervention, policy, or social change may raise questions that previous research could not have answered.
You may also encounter a claim that is widely accepted despite weak evidence, a finding that has not been adequately tested under relevant conditions, or an important population whose circumstances are poorly represented in existing evidence.
The common feature is uncertainty. If reliable evidence already provides a sufficient answer for the purpose at hand, simply repeating the question may contribute little unless there is a defensible reason for replication, verification, extension, or re-examination.
Use the Literature to Establish the Problem, Not Merely to Find an Empty Space
Identifying a research problem requires familiarity with what is already known. Preliminary searching can reveal major findings, disagreements, methodological limitations, theoretical debates, changing conditions, and questions that previous researchers have been unable to resolve.
As your idea develops, the literature review becomes more important because claims about what is unknown must be supported. You cannot confidently call something a research gap after reading only a few convenient papers.
Look across relevant studies rather than relying on one article's “future research” section. Authors often recommend extensions of their own work, but a suggested future direction is an invitation to investigate further, not proof that the proposed study would be important.
A Literature Gap and a Research Problem Are Not Identical
This distinction can prevent a great deal of weak research.
Gap
Something appears to be missing, limited, inconsistent, or insufficiently examined in the existing literature.
Problem
The missing or inadequate knowledge has consequences that give researchers a defensible reason to investigate it.
Suppose you discover that a particular phenomenon has never been studied in one specific city. That is an absence in the literature. It becomes a stronger research problem only if there is a reason the city or its population changes what we need to know. If the location adds no theoretically, practically, methodologically, or contextually meaningful difference, the geographic gap alone may be weak justification.
Ask not only, “What has nobody studied?” but also, “What do we still need to understand?”
Apply the “So What?” Test
Once you can state what is uncertain, ask what changes if the uncertainty is resolved.
Would the answer refine or challenge a theory? Improve a method? Help researchers interpret an important phenomenon? Inform professional practice? Improve decision-making? Provide evidence relevant to policy? Establish whether a previous finding is robust? Reveal whether an assumption is wrong? Provide a foundation for more consequential research?
The contribution does not have to transform an entire discipline. Many worthwhile studies make narrower contributions. What matters is that you can explain a credible pathway from answering the question to improving knowledge or decisions.
The “So What?” test
We do not know X.
Why do we need to know X?
Previous studies disagree about X.
What uncertainty does that disagreement create, and for whom?
X has not been studied in this population.
Why might this population produce meaningfully different evidence?
A previous study had limitation X.
Could correcting that limitation materially change what we conclude?
Answering the question would change little.
Reconsider whether this is the strongest problem to pursue.
Significance Does Not Require a Dramatic Breakthrough
Researchers sometimes reject reasonable ideas because they believe a worthwhile problem must be revolutionary. That sets the wrong standard.
A study can contribute by confirming, refuting, refining, or extending previous findings. Replication can be valuable when there is a substantive reason to test the robustness or generalizability of an important result. Research may also improve measurement, clarify a mechanism, resolve contradictory evidence, document an important phenomenon, or generate knowledge necessary for subsequent studies.
The scale of the expected contribution should be appropriate to the project. An undergraduate project, master's thesis, doctoral dissertation, and large funded research program do not need to solve problems of identical magnitude.
Separate Scientific Importance From Personal Interest
Your curiosity is valuable. It can help you notice problems and sustain your motivation through a difficult project. But personal interest and research significance are not interchangeable.
A question can fascinate you without making a meaningful contribution to the field. Conversely, an important problem may be unsuitable for you if you lack the interest, expertise, access, or resources required to investigate it properly.
A strong project usually needs both: a problem that warrants investigation and a researcher or team capable of pursuing it seriously.
Check Whether Research Can Actually Address the Problem
Some problems are important but cannot be answered by the study you are proposing. Others are framed as research problems even though they are primarily administrative, political, moral, engineering, or operational decisions.
Ask what evidence would reduce the uncertainty you have identified. Can relevant concepts be investigated using appropriate quantitative, qualitative, mixed, theoretical, computational, historical, or other disciplinary methods? Can the required evidence be obtained and interpreted credibly?
If you cannot describe what evidence could move you toward an answer, the problem may still be too vague or may need to be reformulated.
Evaluate Feasibility Before Falling in Love With the Problem
A problem can be important and still be a poor choice for your particular project. Feasibility depends on several critical factors, including time constraints, data accessibility, sample size, funding, equipment, technical expertise, institutional support, and a manageable scope.
Feasibility should be considered early because discovering halfway through a project that the necessary evidence cannot be obtained wastes time and resources. In some cases, the problem remains worthwhile but the proposed study needs to become smaller, use a different method, or form one stage of a larger research program.
Use FINER as a Stress Test, Not a Substitute for Judgment
The FINER criteria provide a useful framework for evaluating an emerging research question: Feasible, Interesting, Novel, Ethical, and Relevant. They can also help you think critically about the research problem from which that question develops.
| Criterion |
Question to Ask |
Why It Matters |
| Feasible |
Can this problem be investigated with the available time, expertise, participants, data, funding, and other resources? |
An important problem does not automatically produce an achievable study. |
| Interesting |
Will the answer sustain your interest and matter to relevant researchers or stakeholders? |
Research requires sustained effort and should address more than private curiosity. |
| Novel |
Could the study confirm, refute, extend, refine, or otherwise add meaningfully to existing knowledge? |
Novelty is about contribution, not simply studying something nobody has named before. |
| Ethical |
Can the problem be investigated using an ethically defensible design? |
A valuable question does not justify unethical research. |
| Relevant |
Could answering it advance knowledge, practice, methodology, policy, theory, or future research? |
Relevance provides the reason the answer is worth obtaining. |
FINER is not an automatic scoring system. A problem does not become worthwhile merely because you can write something under each heading. The criteria are prompts for critical judgment, and their interpretation depends on the discipline and research context.
Consider Ethics at the Problem Stage, Not Only During Ethics Review
Ethics is not something to think about only after the methodology has been designed. The value of the knowledge you expect to gain is part of deciding whether the burdens and risks of research can be justified.
If a question is trivial, redundant, or incapable of producing useful knowledge, exposing participants to avoidable burdens or using substantial resources to answer it becomes harder to justify. The ethical requirements differ substantially among types of research, but scientific value and methodological adequacy are closely connected to responsible research planning.
Write a Problem Statement You Can Challenge
Before moving to a formal research question, try expressing the problem in a short working statement. A useful version identifies what is known, what remains uncertain or inadequate, why that uncertainty matters, and what kind of knowledge is needed.
Then challenge every part of it. Is the claimed uncertainty supported by the literature? Is the consequence real or assumed? Would the proposed research actually reduce the uncertainty? Is the problem still current? Is another study already answering it? Could existing data answer it without collecting new data?
If the problem survives those questions, you have a much stronger foundation for developing a focused research question.
