What Are Inclusion and Exclusion Criteria, and How Restrictive Should They Be?

Who Should Qualify for Your Study, and Who Should Not?

Once you have identified the population your research concerns, you still need an operational way to decide whether a particular person or case belongs in the study.

Suppose you are investigating university instructors' experiences with generative AI. Must participants currently be teaching? Should they have used generative AI themselves? For how long? Should part-time instructors qualify? What about instructors who used AI only once?

These decisions become your inclusion and exclusion criteria. They may look like a procedural detail in a methodology section, but they do considerably more than screen participants. Eligibility criteria help define the population represented by the study, influence recruitment feasibility, and affect how broadly the findings may apply.

The challenge is not to create the longest possible list of requirements. It is to identify the criteria that are genuinely necessary for answering the research question safely, ethically, and meaningfully.

Eligibility Criteria Should Be Necessary, Explicit, and Justifiable

How to Set Defensible Eligibility Criteria

What Are Inclusion Criteria?

Inclusion criteria are characteristics or conditions that must be satisfied for a person, case, record, document, or other unit to be eligible for the study.

In human-participant research, these criteria might concern age, occupation, enrollment status, geographic location, diagnosis, exposure to a particular experience, length of employment, language ability when methodologically necessary, or another characteristic directly relevant to the research question.

The National Institutes of Health describes inclusion criteria in clinical research as factors that allow a person to participate and exclusion criteria as factors that prevent participation. Although the specific examples in clinical research often involve characteristics such as disease status, treatment history, or health conditions, the underlying logic is applicable more broadly: eligibility rules operationalize who belongs in the study.

For example, if you are investigating how university instructors use generative AI when designing assessments, an inclusion criterion might require participants to be currently teaching a higher-education course and to have personally used a generative AI system for an assessment-related activity.

The second condition matters because the research concerns actual experience rather than opinions about a technology participants have never used.

What Are Exclusion Criteria?

Exclusion criteria identify conditions under which a person or case should not participate or should not be included in the study.

An exclusion criterion should ordinarily have a substantive purpose. In some research, exclusion is necessary for participant safety. In other studies, it may prevent the inclusion of cases that cannot answer the research question, avoid serious confounding under a particular design, satisfy ethical or regulatory requirements, or ensure that measurements are interpretable.

Suppose a study examines the experiences of instructors who independently decide whether to use generative AI in their courses. Instructors whose AI use is entirely prescribed by a separate experimental intervention might reasonably be excluded if that circumstance changes the phenomenon being studied.

Exclusion should not simply become a second list of the logical opposites of every inclusion criterion.

Do You Need Both an Inclusion and Exclusion Version of Every Criterion?

No. This often creates redundant methodology.

If an inclusion criterion states that participants must be currently employed as university instructors, you generally do not need a separate exclusion criterion stating that people who are not currently employed as university instructors will be excluded. They already fail the inclusion criterion.

A useful exclusion criterion adds information that is not adequately captured by the inclusion requirements.

For instance, participants might satisfy the basic population definition but have another characteristic that makes their participation inappropriate for the specific design. That characteristic can then be stated explicitly as an exclusion criterion.

Where Should Eligibility Criteria Come From?

Each criterion should be traceable to a methodological reason. Ask why the restriction exists and what would happen if you removed it.

This logic is consistent with the NIH principle of fair participant selection in clinical research: recruitment should primarily follow the scientific goals of the study rather than unrelated characteristics, and groups should not be excluded without an appropriate scientific or risk-based reason.

How Restrictive Should the Criteria Be?

There is no universally correct number of eligibility criteria. The better question is whether every restriction earns its place.

Criteria that are too broad can introduce participants or cases that do not meaningfully correspond to the research question. Criteria that are unnecessarily narrow can create a highly selected group that differs substantially from the population the research is supposed to inform.

Consider a study about generative AI adoption among university instructors. Requiring participants to be current university instructors may be essential. Requiring at least some relevant experience with generative AI may also be necessary if the study concerns actual use.

But suppose the researcher additionally requires participants to be full-time, between 30 and 50 years old, have at least five years of teaching experience, teach face-to-face courses, hold a master's degree, and work in one particular discipline. Unless each restriction follows from the research question or design, the criteria may be manufacturing a much narrower population than intended.

Should Age, Sex, Race, or Other Demographic Characteristics Be Used as Eligibility Criteria?

Only when their use is justified by the research question, safety, ethics, applicable policy, or another legitimate methodological reason.

Unnecessary demographic restrictions deserve particular scrutiny because they can systematically remove groups from the evidence base. In NIH-funded clinical research, for example, current policy expects women and members of racial and ethnic minority groups to be included unless a clear and compelling justification establishes that exclusion is inappropriate to the participants' health or the research purpose. NIH application guidance similarly expects age-based exclusions in relevant clinical research to have a scientific or ethical rationale.

Those requirements are specific to the research governed by those policies, rather than universal rules for every study. The broader methodological lesson is still useful: demographic exclusions should not be inserted reflexively merely because similar criteria appeared in another study.

Criteria Must Be Operational Enough to Apply Consistently

“Experienced teacher” is not a particularly useful eligibility criterion unless experienced is operationally defined. Neither is “frequent AI user” unless the study specifies what counts as frequent use.

Criteria should allow researchers applying the protocol to reach consistent eligibility decisions. If two members of the research team could reasonably classify the same participant differently because a criterion is vague, the criterion needs refinement.

This is one reason eligibility decisions should be established before recruitment rather than improvised as potential participants appear.

Eligibility Criteria and Sampling Method Are Different Decisions

Eligibility tells you who may participate. Sampling tells you how eligible people or cases are selected or recruited.

A study can have carefully specified eligibility criteria and still use convenience sampling. Another study may apply the same criteria and select eligible participants through a probability-based procedure. Eligibility alone does not make a sample representative or unbiased.

After establishing who qualifies, you still need to determine whether a probability or non-probability sampling approach fits your research objective.

Eligibility Criteria Should Be Reported Transparently

Readers need to know how participants entered the study in order to interpret the findings. Reporting guidelines reflect this principle. The STROBE statement for observational studies, for example, asks authors to report eligibility criteria together with the sources and methods used to select participants.

Do not reduce the methodology to “participants who met the inclusion criteria were recruited” without stating what those criteria actually were. The criteria are part of the design, not an administrative detail hidden behind recruitment.

From a Research Question to Defensible Eligibility Criteria

The important step is not merely writing criteria. It is challenging each one: What methodological purpose does this restriction serve? If the only answer is “It makes the sample more specific,” that may not be sufficient justification.

Common Mistakes When Setting Inclusion and Exclusion Criteria

Make Every Eligibility Rule Earn Its Place

Draft your criteria only after the research question and population are clear. Then interrogate each proposed restriction rather than assuming specificity is inherently desirable.

Once the criteria are finalized, examine how they have changed the population to which your findings may apply. Eligibility decisions and external validity are not separate conversations. One helps determine the other.

Before You Finalize Your Eligibility Criteria

Questions About Inclusion and Exclusion Criteria

Restrict Eligibility Only as Much as the Study Requires

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.