How Do You Define the Target Population for Your Research?

Who Exactly Are You Trying to Learn About?

Suppose your research examines how generative AI affects students' academic writing. You describe your target population as university students.

But which university students? Undergraduates or postgraduates? Students enrolled in writing-intensive courses? Students who have used generative AI, or all students regardless of prior experience? Students attending particular institutions, or students across an entire national higher education system?

Each answer changes the population to which your findings are intended to apply.

A similar problem arises in clinical research when investigators describe their population simply as patients, or in organizational research when employees are treated as though they form one interchangeable group.

Defining the target population is therefore not a formality completed after choosing participants. It is a substantive research design decision that establishes the intended scope of inference and influences eligibility criteria, sampling, measurement, and interpretation.

Define the Target Population Through the Research Question and Relevant Boundaries

How to Specify a Target Population That Matches Your Research Question

What Is a Target Population?

The target population is the collection of entities about which a study intends to make descriptive, associational, predictive, causal, or other specified claims.

These entities may be individuals, households, classrooms, schools, hospitals, organizations, communities, countries, or other analytical units.

For example, a study investigating the prevalence of generative AI use among undergraduate students may target all eligible undergraduate students enrolled in a defined higher education system during a particular academic period.

A study examining organizational adoption of AI policies may instead target universities as institutions rather than the students or employees within them.

The target population should therefore correspond to the level and purpose of the research question.

It is not necessarily identical to the people who eventually participate or the records available for analysis.

Why Must the Target Population Be Defined Before Sampling?

Sampling is meaningful only in relation to a population that the researcher intends to represent or investigate.

Without a clear population definition, researchers cannot adequately assess whether their sampling frame covers the relevant entities, whether eligibility criteria are appropriate, or whether selected participants reflect the intended scope.

For example, a sample of 500 students may be appropriate for studying students enrolled in one university but insufficiently representative of students across a national higher education system.

The difference is not merely the number of participants. It concerns which institutions and students can enter the study and what population claims the design can support.

Researchers should establish the distinction between the target and accessible populations before finalizing recruitment procedures.

What Information Should a Target Population Definition Include?

A useful definition identifies the characteristics necessary to determine whether an entity belongs to the population.

The relevant characteristics depend on the research question and may include the unit of interest, educational or clinical status, setting, geography, time period, and substantive experience.

Not every study requires every dimension. The definition should include characteristics that materially determine membership and interpretation.

Adding restrictions simply because they are easy to verify can unnecessarily narrow the population.

How Do You Begin With the Research Question?

Start by identifying the subject of the principal claim.

Consider the question: What proportion of undergraduate students use generative AI for academic writing?

The outcome is a proportion of students, so the target population should include students whose use or nonuse contributes to that proportion.

If the study instead asks how experienced users evaluate the usefulness of generative AI, restricting the population to students with relevant experience may be appropriate.

The same restriction would be inappropriate for estimating the prevalence of AI use because excluding nonusers would remove part of the denominator.

Thus, population membership should be determined by the estimand, not by the researcher's preference for participants who can provide more detailed responses.

How Specific Should the Population Definition Be?

The definition should be precise enough to establish who belongs to the population without introducing unnecessary limitations.

For example, undergraduate students may be too broad if the study specifically examines clinical placements among nursing students.

Conversely, restricting a study of general undergraduate AI use to third-year students from one department may be unnecessarily narrow unless the research question requires that restriction.

Researchers should distinguish substantive eligibility requirements from operational recruitment constraints.

A population can be broadly defined even when only a narrower accessible population can realistically be recruited, provided the resulting limitations on inference are acknowledged.

The appropriate breadth of a population definition is considered further in determining how broad a target population should be.

Should Geography Be Included?

Geography should be included when location is part of the intended population or when geographic context materially affects the research question.

For example, a study of public school teachers in the Philippines has a different population from a study of teachers in Metro Manila.

Geographic boundaries may reflect educational systems, healthcare arrangements, regulatory contexts, cultural settings, or other substantive differences.

However, researchers should not automatically define the target population as the location where data collection happens.

Collecting data in one city does not necessarily mean that the research question is inherently limited to that city. It may instead indicate a narrower accessible population or a particular study setting.

The distinction matters when evaluating whether geography belongs in the target population definition.

Should the Time Period Be Specified?

Yes, when population membership or relevant characteristics change over time.

For example, students enrolled during one academic year may differ from students enrolled several years later.

Similarly, a study of teachers using generative AI in 2026 concerns a technological and institutional context that may change substantially in subsequent years.

Specifying a reference period helps establish which entities are eligible and how the findings should be interpreted.

The period may be a semester, academic year, calendar year, recruitment interval, or another substantively appropriate window.

Researchers should distinguish the population reference period from the dates on which data happen to be collected.

Should Exposure or Experience Be Part of the Definition?

Only when the research question requires it.

For example, a qualitative study examining how experienced researchers negotiate AI authorship decisions may reasonably target researchers who have encountered those decisions.

By contrast, a survey estimating the prevalence of AI authorship practices should not restrict eligibility to researchers who already report engaging in those practices.

Doing so would exclude individuals necessary to estimate the prevalence accurately.

The same principle applies to clinical exposures, professional experiences, and educational interventions.

Population membership should not be defined using the outcome in a way that prevents the intended comparison or distorts the denominator.

How Does the Target Population Differ From the Accessible Population?

The target population consists of the entities about which the researcher intends to make claims.

The accessible population consists of entities that can realistically be reached or identified through the study's recruitment arrangements.

For example, a researcher may target all undergraduate students in a national higher education system but have access only to students enrolled in five cooperating universities.

The accessible population is therefore narrower than the target population.

This difference does not automatically make the research invalid. However, it limits the strength of population-wide inference unless the design and additional evidence support broader generalization.

Researchers should distinguish target population, accessible population, and study sample rather than treating them as equivalent.

How Is the Target Population Different From the Sampling Frame?

The sampling frame is the operational mechanism used to identify units available for selection.

For example, an official enrollment register may serve as the sampling frame for a university student survey.

However, the register may omit eligible students, contain outdated records, or include duplicate entries.

The target population remains the substantively defined collection of eligible students, while the frame determines who can actually enter the selection process.

Researchers should evaluate whether the sampling frame matches the intended population before implementing probability sampling.

A complete frame is valuable, but its completeness must be assessed against the correct population definition.

How Do Inclusion and Exclusion Criteria Relate to the Target Population?

Eligibility criteria operationalize which entities may enter a particular study.

Some criteria define the target population itself. Others reflect safety, measurement, feasibility, or design considerations that restrict the study population more narrowly.

For example, requiring current undergraduate enrollment may be fundamental to a study of undergraduate students.

Requiring students to own a personal laptop may instead be an operational restriction that excludes students whose experiences are relevant to the research question.

Researchers should distinguish scientifically necessary eligibility criteria from avoidable restrictions introduced for convenience.

The implications of inclusion and exclusion criteria for generalizability should be considered before recruitment.

Can the Target Population Be Defined Too Broadly?

Yes. An excessively broad definition may combine entities whose circumstances, exposures, or outcomes differ in ways that make the research question difficult to interpret.

For example, studying the effects of an educational intervention among all learners may be insufficiently specific if the intervention is designed for undergraduate mathematics courses.

Broad populations can also create practical problems when the available sampling frame covers only a small portion of the intended entities.

However, population heterogeneity is not inherently undesirable. It may be important for estimating effects across diverse contexts or examining variation between subgroups.

The issue is whether the research design can meaningfully address the population as defined.

Can the Target Population Be Defined Too Narrowly?

Yes. Unnecessary restrictions can reduce the relevance and transportability of findings.

For example, a study of educational technology use may exclude students with limited internet access because online recruitment is more convenient.

Yet those students may be particularly relevant to understanding digital learning inequalities.

Restricting the population to highly connected students may simplify data collection while changing the phenomenon being investigated.

Researchers should justify narrow definitions according to the research question rather than the convenience of obtaining participants.

A narrowly defined population can be appropriate when the scientific objective genuinely concerns a specific subgroup.

What Is the Difference Between a Population for Description and a Population for Causal Inference?

Descriptive studies often aim to estimate a population characteristic, such as the prevalence of AI use among students.

Causal studies aim to estimate an intervention or exposure effect for a specified population under relevant assumptions.

The populations may differ because causal inference requires attention to treatment eligibility, intervention feasibility, comparison conditions, and potential effect heterogeneity.

For example, a trial of an AI tutoring intervention might target students enrolled in courses where the intervention can realistically be implemented.

Its treatment-effect estimand may not automatically apply to students in courses with substantially different instructional conditions.

Researchers should specify the population relevant to the intended causal contrast rather than assuming that all members of a broad educational category are equally represented.

Does Every Study Need a Statistically Representative Target Population?

No. Research purposes differ.

Probability-based population surveys may require sampling designs that support population estimation.

Qualitative studies may instead define populations or participant groups according to experiences, roles, or circumstances relevant to conceptual understanding.

Experimental studies may prioritize a credible causal comparison within a defined eligible population.

Prediction studies may focus on the population in which a model will eventually be applied.

The target population should reflect the study's inferential purpose rather than an assumption that every research design must produce statistically representative population estimates.

Can a Target Population Include Institutions Rather Than People?

Yes. The target population may consist of schools, hospitals, organizations, communities, or other collective entities.

For example, a study of institutional AI governance policies might target recognized higher education institutions within a specified jurisdiction.

Individual administrators may provide information about those institutions, but the principal analytical claims concern institutions.

Researchers should distinguish the target population from the people recruited as informants.

This distinction is particularly important when individual-level data are used to answer group-level research questions.

Can the Target Population Change During the Study?

Population definitions may occasionally require revision because of new scientific information, operational constraints, or changes in the research question.

However, changing the target population after examining results can introduce concerns about selective interpretation.

Researchers should document substantive changes, explain their rationale, and evaluate how they affect sampling, eligibility, and inference.

When the population definition changes materially, the estimand may also change.

For example, narrowing a study from all university students to students who regularly use generative AI changes the denominator and potentially the research question.

Population revisions should therefore be treated as methodological decisions rather than editorial adjustments.

How Can You Write a Clear Target Population Statement?

A practical statement should identify the entity, relevant substantive characteristics, geographic or institutional scope, and reference period.

For example:

This statement identifies the intended units, educational status, institutional scope, geography, and time period.

It also makes clear that prior AI use is not a membership requirement, which would be important if the research aims to estimate the prevalence of AI use.

The wording does not establish that the researcher can access or sample all such students. That question must be addressed separately through the sampling design.

How Should the Target Population Be Reported?

The methods section should provide an explicit target population definition and explain why its boundaries match the research question.

Describe relevant geographic, institutional, temporal, and substantive eligibility characteristics.

Then distinguish the target population from the accessible population, sampling frame, selected sample, and analytic sample.

Where restrictions arise from feasibility rather than the research question, identify them as limitations rather than silently incorporating them into the original target population.

For observational studies, STROBE emphasizes transparent reporting of setting, participants, eligibility criteria, and selection procedures.

Clear population reporting enables readers to evaluate the relevance and generalizability of the findings.

Defining the Population for a Study of Generative AI Use Among University Students

A defensible target population statement might be:

The distinctions reveal a potential mismatch between the intended national population and the available institutional coverage.

If the five universities were selected conveniently, a large student sample would not automatically support a nationally representative prevalence estimate.

The researcher could consider a broader probability-based institutional sampling design, restrict the population claim, or present the results as evidence from the participating institutions with appropriate limitations.

The population definition helps identify these design decisions before the analysis rather than after the findings are known.

Common Mistakes When Defining a Target Population

How to Define a Target Population Before Choosing Participants

Begin with the intended claim and identify the entities to which that claim refers.

Then specify only the population characteristics necessary to establish meaningful membership and support the research question.

A good population definition is neither the broadest imaginable category nor the smallest group easiest to recruit. It is the collection of entities that corresponds most defensibly to the research question and intended inference.

What to Verify Before Finalizing Your Target Population

Frequently Asked Questions About Defining Target Populations

Define the Population Your Research Question Actually Concerns

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.