How Do You Decide Which Part of a Broad Topic Is Most Worth Studying?

Which Part of a Broad Topic Deserves Your Attention?

A broad topic can contain dozens of possible studies. If you are interested in generative AI in higher education, for example, you could investigate learning, assessment, academic integrity, faculty practice, student attitudes, institutional policy, accessibility, feedback, disciplinary differences, or something else entirely.

Narrowing the topic is therefore not just a matter of making it smaller. You also have to decide which smaller part is actually worth investigating.

That decision becomes difficult because several criteria can point in different directions. One problem may be scientifically important but difficult to study. Another may be feasible but already well understood. A third may be relatively unexplored, yet there may be a good reason researchers have paid little attention to it.

The task is to identify a researchable problem whose answer could make a defensible contribution, rather than simply finding an empty corner of the literature.

Look for the Strongest Intersection of Importance, Uncertainty, and Feasibility

A Worthwhile Research Focus Has to Do More Than Look Unexplored

Researchers sometimes approach a broad topic as though the objective were to locate whichever subtopic has received the least attention. That can be useful for generating possibilities, but publication volume alone does not establish research value.

A topic can be understudied because evidence is genuinely missing. It can also be understudied because the question has little theoretical or practical significance, because relevant data are difficult to obtain, or because another body of literature already addresses the underlying issue using different terminology.

Choosing what to study therefore requires judgment about both the state of knowledge and the value of reducing a particular uncertainty.

Start With a Problem, Not an Empty Space

A literature gap becomes interesting when the missing knowledge creates a problem for understanding, explanation, decision-making, policy, practice, theory, or future research.

For example, discovering that few studies have examined generative AI use among students in a particular province establishes scarcity in a narrow sense. It does not yet establish why another geographically bounded study is needed. The stronger question is whether something about that population or setting creates a meaningful uncertainty that existing evidence cannot adequately resolve.

The difference is subtle but consequential. "Nobody has studied this exact combination" is an observation about the literature. "We do not know whether an established finding applies under conditions that could plausibly change it" begins to articulate a research problem.

Ask What We Still Cannot Explain, Determine, or Decide

Useful opportunities often appear where existing evidence is incomplete in a consequential way. Studies may disagree. An important population may be poorly represented. A widely used intervention may have uncertain outcomes. A theory may make competing predictions. An emerging phenomenon may create questions that earlier research could not have addressed.

In other cases, the contribution may involve testing whether established findings hold under materially different conditions, improving measurement, examining mechanisms, or bringing previously disconnected evidence together.

These possibilities are more informative than treating "the gap" as a single category. What matters is the nature of the uncertainty and what becomes possible if it is reduced.

Importance Is Not the Same as Size

A research problem does not have to concern millions of people to be important. A focused problem affecting a relatively small population may have substantial scientific, ethical, educational, clinical, or policy significance.

Conversely, attaching a study to a large social issue does not automatically make the specific research question important. "Climate change," "artificial intelligence," "mental health," and "educational inequality" are important broad subjects, but the significance of an individual study depends on the particular question it asks and the contribution its answer could make.

This is why narrowing a broad topic should preserve the part of the problem that makes the investigation consequential.

Novelty Does Not Require a Completely Unstudied Topic

The FINER criteria provide a useful framework for evaluating research questions: feasible, interesting, novel, ethical, and relevant. Within this framework, novelty does not have to mean that nobody has investigated anything similar. Research can contribute by confirming, refuting, or extending previous findings.

That matters because searching exclusively for untouched topics can push researchers toward increasingly obscure questions. Replication, extension, testing in a meaningfully different population, improved measurement, or resolving inconsistent evidence may sometimes offer a stronger contribution than studying something merely because it appears new.

Relevance Asks Who Could Use or Build on the Answer

One useful test is to imagine that the study has been completed successfully. Who would care about the answer, and why?

The answer might matter to other researchers because it changes how a phenomenon is understood. It might inform practitioners, institutions, communities, policymakers, or other stakeholders. It may clarify whether a theory applies under particular conditions or establish evidence needed before a larger study becomes sensible.

Not every study needs immediate practical application. Basic and theoretical research can be highly consequential. The point is that you should be able to articulate what kind of knowledge the study would add and why that knowledge deserves attention.

Feasibility Can Eliminate Otherwise Excellent Ideas

A scientifically compelling question may still be unsuitable for your particular project. Feasibility includes access to participants or data, appropriate expertise, time, funding, equipment, measurement, sample size, institutional support, and a design capable of answering the question.

This constraint is especially important for student projects and other research conducted within fixed deadlines. A question that requires a five-year follow-up is unlikely to become feasible merely because the thesis deadline approaches with scholarly determination.

Feasibility should therefore be assessed early, but it should not become the sole criterion. Choosing a question only because the data are easy to collect can produce a technically manageable study with little reason to exist. When scientific importance and practicality conflict, consider explicitly how much feasibility should influence the scope.

Ethical Acceptability Is a Requirement, Not a Competitive Advantage

A potentially informative study is not worth conducting through ethically unacceptable means. Ethical considerations may affect recruitment, intervention, data collection, privacy, risk, consent, inclusion, and the populations that can appropriately be studied.

Ethics can also affect the value of the research question itself. Research involving people should justify the burdens and risks imposed on participants through a reasonable prospect of producing useful knowledge.

Use Several Criteria Together Rather Than Searching for One Winning Criterion

These criteria are not a mechanical scoring system. A study with moderate novelty but substantial relevance may be more worthwhile than a highly novel question with little significance. Likewise, an important problem may need to be reformulated rather than abandoned when the first version is infeasible.

The objective is to find a defensible combination, not to maximize every criterion independently.

From a Broad Interest to a Study Worth Conducting

This does not prove that the selected topic is objectively the best possible study. Another researcher with different expertise, participants, resources, or theoretical interests might reasonably choose another part of the same broad topic. A single research area can support multiple distinct and worthwhile studies.

Common Ways Researchers Mistake Novelty for Research Value

Compare Candidate Problems Before Committing to One

If your broad topic contains several possibilities, do not immediately ask which one sounds most impressive. Write each possibility as a provisional research problem and interrogate it.

For each candidate, identify what is currently uncertain, what evidence supports the existence of that uncertainty, why resolving it could matter, and what study would actually be required. This often exposes the difference between an interesting subject and an investigable problem.

You may discover that two possibilities remain similarly compelling. At that point, the problem is no longer simply how to narrow. You need to choose between competing research topics using criteria that matter for your particular project.

Keep the decision provisional while you investigate the literature. A promising idea can become weaker once you discover what has already been established, while an apparently modest question may become more consequential once you understand the unresolved debate around it.

Check Whether This Part of the Topic Is Worth Pursuing

Questions About Choosing the Most Worthwhile Research Focus

Choose the Problem Whose Answer Would Actually Add Something

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.