01 · The Question
Every Study Draws a Boundary Around Who Counts
Every study involving people makes decisions about who can participate. Some of those decisions are explicit: an age range, a diagnosis, a location, a particular occupation, or prior exposure to an intervention. Others are less obvious. Requiring participants to travel to a research site during working hours, complete a survey in only one language, use a smartphone, or respond to an online recruitment advertisement can also determine who ultimately appears in the study.
This matters because participant selection is not simply a logistical step between designing a study and collecting data. The people who are eligible, reachable, willing, and able to participate become the empirical basis for the conclusions researchers draw. People who are systematically left out may consequently be absent from the evidence used to understand a problem, evaluate an intervention, or inform practice.
The difficult question is therefore not simply, “Who should be in my sample?” It is also: Who has a reasonable opportunity to be included, who does not, and can those boundaries be justified by the research question, study design, and ethical requirements?
03 · What You Need to Know
Participant Selection Happens at More Than One Point in a Study
It is tempting to think that inclusion and exclusion are settled when the protocol lists eligibility criteria. In practice, the eventual sample is shaped by a sequence of decisions. A person may belong to the population of interest yet never encounter the recruitment message. Another may see it but fail an eligibility requirement. Someone else may qualify but be unable to attend the required sessions. A fourth person may enroll but withdraw because participation becomes impractical.
Understanding who gets studied therefore requires looking beyond the final participant list.
Start With the Population the Research Question Is About
The first boundary comes from the research question itself. A study does not ordinarily need to include everyone. If researchers are investigating first-year university students' adjustment to university life, limiting participation to first-year students is substantively connected to the question. If a study evaluates an intervention intended specifically for people with a particular clinical condition, requiring that condition may likewise be necessary.
This is different from excluding people merely because including them would make recruitment or data collection more difficult.
The key principle is alignment: the study population should correspond to the population needed to answer the research question. This also means that broader inclusion is not automatically better. A population can become so broadly defined that participants no longer share the characteristics necessary for a meaningful answer. The challenge is to balance inclusion with a clearly defined study population.
Formal Inclusion and Exclusion Criteria Create the Eligibility Boundary
Inclusion criteria identify characteristics a person must have to participate. Exclusion criteria identify circumstances that make an otherwise relevant person ineligible. Depending on the study, criteria might concern age, diagnosis, prior treatment, employment status, educational level, geographic location, ability to provide informed consent, or other characteristics directly relevant to the design.
These criteria can protect participants, reduce confounding, define the intended population, or ensure that the study can answer a sufficiently specific question. They are not inherently problematic.
The problem arises when a criterion has little scientific or ethical justification yet systematically removes people who are part of the population the research is meant to inform. In such cases, eligibility criteria can contribute to an unrepresentative sample and narrow what the resulting evidence can support.
Eligibility Is Not the Same as a Real Opportunity to Participate
A person can satisfy every inclusion criterion and still be effectively excluded.
Imagine a study that technically permits working adults to participate but schedules all interviews between 10:00 a.m. and 3:00 p.m. on weekdays. Or a community survey open to older adults that can be completed only through a QR code and smartphone interface. Neither design formally excludes people who work during the day or those with limited access to digital technology. Yet the procedures may make participation substantially harder for them.
Practical barriers can arise from transportation, disability access, caregiving responsibilities, internet connectivity, technology requirements, compensation arrangements, literacy demands, scheduling, recruitment channels, or the time required to participate. Researchers therefore need to consider how study procedures themselves may unintentionally exclude participants.
Language Can Function as an Eligibility Criterion Even When It Is Not Written as One
Language illustrates how an operational decision can quietly become a population boundary. A study may say that adults in a particular community are eligible, while recruitment materials, consent documents, surveys, interviews, and participant support are available in only one language.
Sometimes a language restriction is methodologically defensible. A study may investigate a language-specific phenomenon, for example, or validated instruments may not yet be available in other languages. In other cases, the restriction reflects available resources rather than the phenomenon being studied.
Researchers should distinguish those situations because language requirements can exclude populations whose experiences are relevant to the research question.
Recruitment Determines Who Even Gets the Chance to Volunteer
Eligibility criteria answer, “Who may participate?” Recruitment adds another question: “Who will actually hear about the study?”
Recruiting only through one university, hospital, online platform, professional association, neighborhood, or social network can produce a much narrower pool than the nominal target population. Convenience sampling is sometimes appropriate, particularly for exploratory research or studies with deliberately bounded populations. The limitation should nevertheless be recognized rather than allowing the accessible population to quietly stand in for a much broader one.
The Belmont Report's principle of justice is particularly relevant here. It links justice to fair procedures and outcomes in participant selection and cautions against selecting people for risky research merely because they are readily available or disadvantaged. The same principle also directs attention to the distribution of research benefits, not only its burdens.
Protection and Exclusion Are Not the Same Thing
Some populations require additional ethical safeguards. That does not necessarily mean they should automatically be absent from research.
Historically, attempts to protect certain populations have sometimes relied heavily on exclusion. International ethical guidance from the Council for International Organizations of Medical Sciences (CIOMS), for example, notes that excluding groups considered vulnerable, including children, women of reproductive age, and pregnant women, has contributed to gaps in evidence relevant to those populations. Contemporary guidance increasingly asks whether participation can be made ethically appropriate rather than assuming that exclusion is always the safest response.
Protection
Identifying relevant risks and using appropriate safeguards, consent processes, monitoring, or study modifications.
Exclusion
Preventing a person or group from participating in the study altogether.
The distinction matters because exclusion can itself have consequences. When a population is repeatedly omitted from studies, clinicians, policymakers, educators, or other decision-makers may eventually have less evidence about whether findings apply to that population.
Inclusion Has Both Scientific and Ethical Dimensions
Participant selection affects what researchers can infer from a study. If people who experience a phenomenon differently are systematically absent, the study may overlook relevant variation. This is one reason researchers should consider why diversity in a research sample may matter for the question being investigated.
Selection is also an ethical issue. The Belmont Report identifies justice as one of the foundational principles of research involving human participants and connects it directly to equitable participant selection. CIOMS similarly emphasizes fair distribution of the benefits and burdens of research. These principles raise two complementary concerns: some groups should not disproportionately bear research burdens merely because they are convenient to recruit, while relevant populations should not be systematically denied opportunities to contribute to and potentially benefit from research without adequate justification.
Watch Out
Do not treat “inclusive” as meaning “include everyone.” Ethical and scientifically sound research can require exclusions. The important question is whether the boundary is relevant, proportionate, and defensible rather than arbitrary or merely convenient.
Being Included Does Not Automatically Mean Being Adequately Represented
Suppose a study enrolls participants from several demographic groups. Technically, those groups are included. That alone does not establish that the sample composition is appropriate for the study's intended inferences.
A group might appear in the dataset but in numbers too small to support meaningful subgroup analysis. Conversely, a sample can include substantial demographic diversity without statistically representing the population from which researchers want to generalize. These are related but distinct concerns, which is why representation and representativeness should not be treated as interchangeable.
Inclusion is therefore best understood as one part of a larger sampling problem. It concerns who has access to participation and whose experiences enter the evidence. It does not, by itself, establish representativeness or guarantee valid generalization.
04 · A Practical Example
How a Broadly Defined Study Can Produce a Narrow Participant Pool
Hypothetical Example
A study of university students' experiences with online learning
Suppose a research team wants to understand online-learning experiences among undergraduate students at a university. The formal inclusion criteria are broad: any currently enrolled undergraduate student aged 18 or older may participate.
Target population All currently enrolled undergraduate students aged 18 or older are relevant to the research question.
Recruitment The invitation is posted primarily on the university's learning management system and student social-media pages.
Participation Students must complete a 40-minute online interview during weekday office hours using a stable video connection.
Result Students with reliable connectivity, flexible schedules, and frequent engagement with university online channels are more likely to encounter the invitation and complete the study.
Interpretation The final participants may satisfy the formal eligibility criteria while still underrepresenting students with poor connectivity, demanding work schedules, caregiving responsibilities, or limited engagement with the recruitment channels.
Action The researchers could examine whether alternative recruitment channels, asynchronous participation options, telephone interviews, or more flexible scheduling are feasible without compromising the study's purpose.
The lesson is not that every possible accommodation must be offered. Rather, researchers should recognize that the realized sample emerges from more than written inclusion and exclusion criteria. Recruitment and participation requirements also filter the population.
06 · What This Means for You
Audit the Path Into Your Study, Not Just the Eligibility List
When designing research involving people, begin with the population required by the research question. Then work through the route a person must travel before becoming a participant. At each point, ask whether a boundary is intentional, necessary, and defensible.
A simple decision framework
If a characteristic is necessary to define the population relevant to the research question
Include it as an eligibility criterion and explain why it is necessary.
If excluding a person or group is necessary for participant safety or another ethical reason
Document the rationale and consider whether appropriate safeguards could address the concern without complete exclusion.
If a restriction exists mainly because it makes recruitment or data collection easier
Ask whether a feasible alternative could reduce the barrier without undermining the study.
If a relevant group is eligible but unlikely to encounter or complete the study
If an important group is likely to remain uncommon in the recruited sample
For research subject to institutional, funder, national, or disciplinary requirements, check the rules that apply to the specific study. Inclusion policies are not universal across all research contexts. For example, current U.S. National Institutes of Health policies establish particular inclusion expectations for NIH-funded clinical research and human-subjects research, including requirements concerning women, racial and ethnic minority groups, and inclusion across the lifespan. Those requirements should not be generalized as though they govern every study everywhere.
More broadly, the aim is not to eliminate every boundary. It is to make the boundaries visible enough that they can be examined. A well-defined study may still be narrow. What matters is that researchers understand why it is narrow, what consequences follow from that decision, and how far the resulting claims can reasonably extend.
07 · A Quick Checklist
Before Finalizing Who Can Participate, Check These Boundaries
Before recruiting participants, check:
Define the population your research question is actually intended to describe, explain, or inform.
Review every inclusion and exclusion criterion and identify its scientific or ethical justification.
Ask whether any criterion exists primarily for researcher convenience and whether a feasible alternative is available.
Examine who will realistically encounter your recruitment materials and who may never see them.
Check whether language, technology, location, scheduling, transportation, disability access, or participation burden creates additional barriers.
Consider whether groups affected by the issue being studied are systematically missing from the likely participant pool.
Distinguish safeguards needed to protect participants from assumptions that a population must be excluded altogether.
Verify applicable ethics, institutional, funder, regulatory, and reporting requirements rather than assuming one inclusion policy applies universally.
Plan to describe important selection limitations when reporting and interpreting the study.
