03 · What You Need to Know
Broadness Is Really a Problem of Scope
A research topic and a research question serve different purposes. A topic identifies an area you want to investigate. A research question defines what, within that area, the study will actually try to answer.
“Social media and mental health” is a topic. “How do first-year university students describe the ways social media use shapes their experiences of academic stress during their transition to university?” is much closer to a research question because it establishes a particular phenomenon, population, context, and form of inquiry.
The transition from a broad interest to an answerable question is a central part of developing a good research question. Narrowing is not simply shortening the topic. It means deciding which part of the larger problem your study can legitimately address.
A broad question contains too many possible studies
One useful diagnostic is to ask how many substantially different studies could claim to answer your question.
Take “How does technology affect student learning?” One researcher could study smartphone distraction among secondary students. Another could examine adaptive tutoring in mathematics. A third could investigate virtual laboratories in engineering education. A fourth might analyze learning management system use among distance learners. All could plausibly say they were studying the original question.
That flexibility may be useful when describing a research area, but it is a warning sign in a question intended to guide one empirical study. The question has not yet established which phenomenon will be investigated.
Look for undefined umbrella concepts
Broad questions often rely on large categories that conceal substantial variation. “Technology,” “AI,” “education,” “health,” “social media,” “leadership,” and “academic performance” can each encompass many different phenomena.
Even apparently specific terms may hide several possible meanings. “Student engagement,” for example, might refer to behavioral participation, emotional engagement, cognitive investment, attendance, interaction with course materials, or another construct depending on the theoretical and methodological context.
The solution is not automatically to replace every broad word with a narrow one. First determine which meaning is relevant to the problem you actually want to investigate. Conceptual precision should drive the wording, not the other way around.
Check whether the population is effectively unlimited
Questions about “students,” “teachers,” “patients,” “employees,” or “consumers” may encompass populations whose experiences and characteristics differ considerably.
Whether a population needs further specification depends on the purpose and design of the study. Age, educational level, occupation, diagnosis, geographical location, institutional setting, or other characteristics may be relevant boundaries. In some research traditions, however, specifying every possible characteristic in the question itself would be unnecessary.
The important distinction is between defining the study and cramming its entire protocol into one sentence. Whether a question should explicitly name the population, variables, context, and outcome depends on the type of research question and disciplinary conventions.
Check how many outcomes or phenomena the question asks you to explain
Scope expands quickly when a single question asks about many outcomes simultaneously.
Consider: “How does social media use affect students' academic performance, mental health, sleep, relationships, self-esteem, physical activity, and career aspirations?” Each outcome could involve different constructs, instruments, mechanisms, confounders, bodies of literature, and analytical decisions.
Combining them does not necessarily produce a more comprehensive study. It may instead produce several partial studies sharing one dataset.
This does not mean a study can have only one outcome or phenomenon. Multiple outcomes may be theoretically coherent and methodologically manageable. The issue is whether they collectively serve one identifiable research problem rather than reflecting an attempt to investigate everything that might be interesting.
Check whether the question spans too many contexts
Context may include institutions, countries, communities, workplaces, healthcare settings, disciplines, educational levels, cultures, or policy environments. A question that silently crosses several of these can become difficult to answer because the phenomenon may operate differently in each setting.
“How do universities use generative AI?” could refer to assessment, teaching, research, administration, student support, institutional policy, academic integrity, or staff development across thousands of institutions operating under different conditions.
You might deliberately study variation across contexts, of course. Comparative and multisite studies do exactly that. In such cases, breadth is part of the design rather than an oversight. The relevant question is whether the chosen contexts are theoretically or practically justified and whether the study has the resources to examine them properly.
Check whether the time frame is doing hidden work
Time can enlarge a question even when it is not explicitly mentioned. “How has online learning affected higher education?” could potentially encompass decades of technological, institutional, pedagogical, and social change.
A time boundary may therefore be needed when the phenomenon changes over time or when historical periods matter to interpretation. In other studies, adding an arbitrary date range merely to make the question look specific would contribute little.
The appropriate level of detail should follow the logic of the study. The separate question of how specific a research question should be before the study begins requires balancing sufficient direction against premature methodological rigidity.
Check whether one method could realistically address the whole question
A question may reveal excessive breadth when different parts require fundamentally different evidence or methods.
Suppose you ask, “How prevalent is academic burnout among university students, what causes it, and how do students experience it?” Estimating prevalence, making causal claims, and understanding lived experience are distinct forms of inquiry. They may be combined within a carefully designed program of research or mixed-methods study, but simply placing them in one sentence does not create methodological coherence.
When different parts of the question demand separate designs, samples, data sources, or analytical traditions, you may actually have multiple research questions. A study can certainly contain more than one research question, but each additional question creates methodological and analytical obligations.
Broad does not mean long
Word count is a poor measure of scope.
Short but extremely broad
“What affects student success?”
Longer but more bounded
“What factors do first-generation undergraduate students identify as shaping their persistence during their first year at a public university?”
The second question contains more words because it establishes boundaries. Those boundaries help determine who or what should be studied and what kind of evidence would be relevant.
A broad question is not automatically a bad question
Broadness must be evaluated relative to the type and purpose of the research.
Exploratory studies may intentionally begin with relatively open questions. Qualitative inquiry can require enough openness for participants' meanings and unanticipated findings to emerge. Scoping reviews also commonly use broader questions than narrowly focused systematic reviews because their purpose may be to map the extent, range, or characteristics of a body of evidence.
Even then, broad does not mean boundaryless. Methodological work on scoping studies has emphasized that broad questions still need an articulated scope, including clarification of concepts and populations, so that searching and study selection remain coherent.
Likewise, qualitative and quantitative research questions can require different kinds of specificity. A useful question is not necessarily the one containing the greatest number of delimiters.
The decisive test is whether the question is manageable and answerable
The FINER criteria describe a strong research question as feasible, interesting, novel, ethical, and relevant. Feasibility includes whether a project is manageable in scope as well as whether the researcher has sufficient participants, expertise, time, and resources.
This gives “too broad” a more useful meaning. The problem is not that a supervisor prefers shorter questions. The problem is that the scope may prevent the study from generating enough appropriate evidence to answer what the question promises.
A broad question can therefore fail the more fundamental test of whether the question is actually researchable. If you cannot identify a realistic study capable of answering the whole question, some part of its scope probably needs to change.
04 · A Practical Example
How a Broad Topic Becomes a Manageable Research Question
Hypothetical Example
From social media and academic performance to one defensible study
A student researcher begins with the question: “How does social media affect the academic performance of students?” The researcher has one semester to complete the study and access to undergraduate students at one university.
Start with the broad question “How does social media affect the academic performance of students?” leaves the platform or type of use, population, outcome, context, and meaning of “affect” insufficiently defined.
Identify the real interest After reviewing the literature, the researcher realizes that the concern is not social media generally but students' use of social media while studying.
Choose the relevant population and context The accessible population is first-year undergraduates enrolled at the researcher's university.
Clarify the outcome Rather than treating “academic performance” as an unlimited concept, the researcher decides that semester grade point average is the outcome relevant to the proposed study.
Match the wording to the feasible evidence The planned observational study can investigate an association, but it cannot by itself establish that social media use causes changes in grades.
Produce a focused question “Among first-year undergraduates at University X, is self-reported social media use while studying associated with semester grade point average?”
The final question is narrower, but narrowing has not made it inherently better. The researcher must still justify the measurement of social media use, consider potential confounding, determine an appropriate sample and analysis, and assess whether self-reported behavior provides sufficiently credible evidence.
What narrowing has accomplished is more modest and more important: it has made the boundaries of the proposed study visible.
