What Questions Remain Unanswered Because Populations Are Too Narrow?

When Does a Narrow Study Population Become a Real Research Gap?

You review the literature and notice that almost every study draws participants from the same kind of population: perhaps university students, urban residents, younger adults, one professional group, people from high-income settings, or participants without common comorbidities.

It is tempting to conclude that every population not represented in those studies constitutes a research gap.

That would go too far.

Research findings do not automatically stop applying when you cross an institutional, geographic, demographic, or cultural boundary. At the same time, evidence generated from a restricted population should not automatically be assumed to apply to everyone else.

The useful question is whether characteristics of the understudied population could plausibly change the phenomenon, effect, mechanism, baseline risk, implementation, measurement, or interpretation. If so, narrow populations may leave an important question genuinely unresolved.

Look for Uncertain Applicability, Not Merely an Unstudied Group

Ask Whether the Evidence Travels to the Population You Care About

Every study has a population boundary

No empirical study includes everyone to whom its findings might eventually be relevant. Researchers define eligibility criteria, recruit from particular settings, obtain participation from some eligible people rather than others, and analyze a finite sample.

Generalization is therefore unavoidable. The important issue is whether the move from the studied population to the target population is defensible.

A narrow population becomes problematic when the evidence is used to support claims about people, settings, or circumstances for which relevant effect-modifying or baseline characteristics may differ.

Representativeness is always representativeness of something

Researchers sometimes describe a sample simply as “representative” or “not representative.” That language can conceal an important question: representative of which target population, for which inference?

A sample of first-year engineering students might be entirely appropriate if the research question concerns first-year engineering students. It becomes narrow only relative to a broader claim, such as one about all university students.

The gap emerges when there is consequential uncertainty in moving from the former to the latter.

Population differences matter when they can change the answer

Not every demographic or contextual difference modifies a finding. A study conducted in one province does not automatically require replication in every other province. Nor does research from one university automatically become invalid outside that institution.

GRADE's treatment of indirectness captures this nuance. Population indirectness becomes a concern when differences between the population represented in the evidence and the population of interest could lead to meaningful differences in relative or absolute effects. GRADE specifically cautions against assuming indirectness merely because populations differ; there should be compelling reasons to expect those differences to matter.

This principle is useful far beyond clinical research.

Ask what mechanism could make the answer different. Age might matter because developmental processes differ. Prior experience might alter how people respond to an educational intervention. Infrastructure might change whether a technology can be implemented. Language could affect comprehension of an instrument. Institutional resources might influence whether an intervention works as intended.

The justification comes from the mechanism, not the demographic label alone.

Baseline conditions can change what an effect means

Even when a relative effect transfers reasonably well across populations, the absolute consequences may differ because baseline conditions differ.

GRADE identifies uncertainty about baseline risk in a target population as an important source of population indirectness. If baseline risk differs, the same relative effect can imply a different absolute benefit or harm.

The broader lesson is that applicability depends not only on whether people look demographically similar. What matters is whether relevant starting conditions, exposures, resources, risks, mechanisms, and contexts differ in ways that alter the interpretation of the result.

Eligibility criteria can systematically remove the people who complicate the question

Researchers often restrict samples for defensible reasons. Narrow eligibility criteria can create a more homogeneous sample, protect participants, reduce confounding, simplify implementation, or focus on a theoretically relevant population.

But those exclusions have consequences.

If studies repeatedly exclude older adults, people with disabilities, individuals with comorbidities, people using other treatments, participants with limited proficiency in the research language, or people facing difficult access conditions, evidence may become strongest precisely for the easiest population to study.

Current NIH policy illustrates the importance of considering this issue explicitly. NIH-supported human-subjects research generally requires inclusion across the lifespan unless scientific or ethical reasons justify age-based exclusion, with the stated aim of generating knowledge applicable to populations affected by the conditions being studied.

Similarly, current FDA guidance on clinical-trial participation encourages broader enrollment practices so study populations more closely reflect people likely to use an approved intervention, including attention to both demographic and non-demographic characteristics.

These are specific U.S. policies for particular research contexts, not universal requirements for all disciplines. The underlying methodological issue, however, is broadly relevant: systematic exclusion can leave applicability uncertain.

Convenience samples are not automatically useless

Convenience sampling receives considerable criticism, sometimes too casually. Whether it is problematic depends on the inference being made.

A convenience sample can provide useful evidence for theory development, exploratory work, measurement research, experimental effects under specified conditions, and many other purposes. The problem arises when researchers make population-level claims that require assumptions the sample cannot support.

Instead of asking whether a sample is convenient, ask whether selection into the study is related to characteristics that matter for the conclusion.

Geographic novelty is usually not enough

“No study has been conducted in Country X” is one of the easiest research gaps to write and one of the easiest to overstate.

A location becomes theoretically or practically important when something about that context could plausibly change the answer.

Underrepresentation and systematic exclusion are related but not identical

A population may be underrepresented because recruitment methods fail to reach it, participation barriers reduce enrollment, eligibility criteria exclude it, the condition is uncommon in that group, or the original research question legitimately concerns another population.

Those explanations should not be treated as interchangeable.

If certain groups are repeatedly excluded despite being directly relevant to the question, the problem may warrant a more focused examination of populations that are systematically left out of research.

Population breadth and population relevance are different

Broader is not always better.

A highly heterogeneous sample may be unnecessary for a tightly defined research question. It can also make interpretation difficult if the study lacks enough information to examine meaningful variation among participants.

The objective is not maximum diversity as an abstract methodological virtue. The study population should correspond to the scientific question and the population to which the conclusions are intended to apply.

NIH's current inclusion guidance makes a similar point in its own context: inclusion is expected in a manner appropriate to the scientific question.

Population gaps should ultimately return to uncertainty

The strongest population-based research gap does not end with “this group has not been studied.” It explains what researchers cannot conclude because the group has not been adequately represented.

Compare these two formulations:

“Few studies have examined AI-assisted learning among working adult students.”

“Existing evidence on AI-assisted learning comes predominantly from traditional full-time students. It remains uncertain whether the same effects occur among working adult students, whose time constraints, prior professional experience, patterns of technology use, and learning conditions may alter how the intervention is used and what benefits it produces.”

The second statement is not automatically correct. Those proposed differences would themselves need evidence or theoretical justification. But it demonstrates the reasoning required: connect population difference to uncertainty about the phenomenon.

When a Large Student Literature Still Has a Population Problem

The distinction matters. The proposed population should be scientifically informative, not simply different.

Common Mistakes When Claiming Population Research Gaps

Justify Why the Missing Population Could Change the Answer

If you identify an understudied population, do not stop at documenting its absence. Build the reasoning that connects population characteristics to the uncertainty you want to resolve.

This approach also helps distinguish a genuine population uncertainty from a generic call for replication. Replication in another population can be valuable, but its justification becomes much stronger when you can state what assumption about generalizability is being tested.

Before Claiming That a Population Is Understudied, Check Why It Matters

Questions About Narrow Study Populations

A Missing Population Matters When Its Absence Leaves Applicability Uncertain

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.