03 · What You Need to Know
How to Develop Inclusion Criteria That Support the Intended Research
What Are Inclusion Criteria?
Inclusion criteria are the conditions an individual or other entity must satisfy to be eligible for a research study.
They may concern population membership, educational or clinical status, age, experience, location, time period, or other characteristics relevant to the research question.
For example, a study of undergraduate academic writing may require current undergraduate enrollment and participation in courses involving relevant writing tasks.
A clinical intervention may require a particular diagnosis and treatment indication.
Inclusion criteria operationalize eligibility. They should not be confused with the sampling method, which determines how eligible units are selected or recruited.
A participant can satisfy every inclusion criterion without necessarily being selected for the sample.
Why Must Inclusion Criteria Follow the Research Question?
The research question determines whose characteristics, experiences, or outcomes are necessary for the intended inference.
Inclusion criteria should preserve that relationship.
For example, estimating the prevalence of AI use requires a population containing both users and nonusers.
By contrast, investigating firsthand experiences of AI-assisted manuscript preparation may reasonably require participants who have engaged in that activity.
Eligibility criteria that contradict the intended estimand can make a study incapable of answering its own question.
Researchers should therefore define the target population before finalizing inclusion criteria.
What Is the Difference Between Population Definition and Inclusion Criteria?
The target population describes the collection of entities about which the study intends to make claims.
Inclusion criteria specify how membership in the eligible study population will be determined operationally.
For example, a target population might be undergraduate students enrolled in recognized higher education institutions during a specified academic year.
The inclusion criteria could require official undergraduate enrollment during that period and membership in an eligible degree program.
However, some operational eligibility restrictions may make the study population narrower than the broader target population.
Researchers should distinguish population-defining criteria from restrictions introduced for safety, measurement, or practical reasons.
What Characteristics Can Be Used as Inclusion Criteria?
Potential criteria depend on the research question and design.
| Criterion type |
Example |
When it may be justified |
| Educational status |
Currently enrolled undergraduate student |
The study concerns undergraduate learning. |
| Professional role |
Currently employed classroom teacher |
The research concerns active teaching practices. |
| Clinical diagnosis |
Confirmed diagnosis using specified criteria |
The research concerns a defined clinical condition. |
| Age |
Within a justified age range |
Age defines the intended population or is required for the intervention. |
| Experience |
Completed at least one relevant AI-assisted writing task |
Firsthand experience is necessary for the question. |
| Time |
Enrolled during a specified semester |
The study concerns that enrollment period. |
| Setting |
Receives care within a defined clinical service |
The service context defines the intended population. |
| Intervention eligibility |
Eligible to receive a specified instructional program |
The study estimates an effect among those for whom the intervention is applicable. |
These examples are not universal templates. The scientific rationale for each criterion must be established within the particular study.
How Do You Translate a Research Question Into Inclusion Criteria?
Begin by identifying the unit about which the research intends to make claims.
Then determine which characteristics are necessary for that unit to belong to the target population.
Next, identify any additional conditions required for the exposure, intervention, outcome, or research procedures to be meaningful.
Finally, translate those conditions into rules that can be assessed consistently before enrollment.
For example, a study of first-year undergraduate students' academic writing may require first-year enrollment and participation in a relevant academic writing course.
It would not automatically require students to have high grades, own personal laptops, or possess previous AI experience unless those characteristics were substantively necessary.
What Makes an Inclusion Criterion Clear?
A clear criterion allows different researchers or recruitment personnel to reach consistent eligibility decisions using the same information.
For example, "must be an active student" may be ambiguous because active enrollment can have different administrative meanings.
A more operational criterion might specify official enrollment in an undergraduate degree program on the institution's designated census date.
Similarly, "must be experienced in AI" is vague unless experience is defined through a relevant activity, frequency, duration, or competence measure.
Clarity requires identifying the characteristic, acceptable evidence, and relevant assessment period.
How Specific Should Inclusion Criteria Be?
Criteria should be sufficiently specific to establish eligibility without introducing unnecessary restrictions.
For example, requiring participants to be enrolled in undergraduate degree programs may be appropriate for a study of undergraduate academic writing.
Requiring enrollment in one particular section may be unnecessary if the research question concerns a broader undergraduate population.
Overly specific criteria can reduce recruitment feasibility and restrict generalizability.
Insufficiently specific criteria can create inconsistent eligibility decisions and include participants for whom the outcome or intervention is not meaningful.
The appropriate level of specificity depends on the intended estimand and research design.
Should Inclusion Criteria Be Measurable?
Yes. Eligibility conditions should be assessable using information that is sufficiently reliable for the study's purpose.
For example, enrollment status may be verified through institutional records, while prior experience may be assessed using a structured screening question.
Clinical eligibility may require diagnostic records, testing, or professional assessment, depending on the condition.
Researchers should specify how each criterion will be evaluated and how uncertain information will be handled.
However, measurement feasibility should not automatically justify excluding otherwise relevant population members.
If a characteristic is important but difficult to verify, researchers should consider improving the assessment method before narrowing the population.
Should Prior Experience Be an Inclusion Criterion?
Only when experience is necessary to answer the research question.
For example, a qualitative study examining experiences with AI-assisted manuscript writing may appropriately require firsthand involvement in that activity.
However, a survey estimating the prevalence of AI use should include both users and nonusers.
Similarly, a study comparing novice and experienced users should not exclude novices.
Researchers should distinguish experience as a defining characteristic from experience as a variable to be measured.
This principle also applies to disease stage, severity, and experience in population definition.
Should Demographic Characteristics Become Inclusion Criteria?
Demographic characteristics should determine eligibility when they are substantively necessary or ethically justified.
For example, a study specifically investigating adolescent experiences may require an age range corresponding to its intended population.
However, restricting a general educational technology study to one age group without justification may exclude relevant participants.
Researchers should consider whether demographic characteristics define the phenomenon, affect intervention eligibility, or should instead be measured for analysis.
Unnecessary restrictions can reduce population coverage and potentially exclude underrepresented groups.
Eligibility criteria should not be used as an informal mechanism for constructing a more convenient or homogeneous sample.
How Do Inclusion Criteria Differ From Exclusion Criteria?
Inclusion criteria identify the conditions necessary for eligibility.
Exclusion criteria identify circumstances that disqualify otherwise eligible units from participation or particular study procedures.
For example, a clinical trial might require a specified diagnosis as an inclusion criterion and exclude individuals with a contraindication to the intervention.
However, the distinction is not always purely linguistic. A criterion can sometimes be expressed positively or negatively without changing the underlying eligibility rule.
The important requirement is that eligibility decisions are coherent, justified, and consistently applied.
Researchers should avoid contradictory or redundant criteria that make recruitment unnecessarily complicated.
Should Every Potential Confounder Become an Inclusion Criterion?
No. Restricting eligibility according to every characteristic associated with the outcome can produce an unnecessarily narrow population.
For example, a study of AI-supported learning should not automatically require students to have similar prior achievement, socioeconomic circumstances, or digital access.
These characteristics may be important variables to measure and address through the research design or analysis.
Restriction can sometimes help control confounding, but it changes the population of inference and may create other selection problems.
Researchers should justify each restriction according to the causal structure and intended estimand rather than using eligibility criteria as a universal confounding-control strategy.
How Can Inclusion Criteria Accidentally Select on the Outcome?
Eligibility criteria can undermine the research question when they require participants to possess the outcome being estimated.
For example, a study estimating the prevalence of AI-assisted academic writing cannot include only students who already use AI for writing.
Similarly, a study estimating the prevalence of academic burnout should not recruit only students who report experiencing burnout.
Such restrictions remove individuals without the outcome from the denominator.
In causal research, conditioning on post-exposure outcomes or related variables can also introduce selection bias.
Researchers should determine whether a proposed criterion defines the population legitimately or improperly conditions participation on the phenomenon under investigation.
What About Inclusion Criteria Based on Future Participation?
Eligibility should generally be determined using information available at the appropriate enrollment or baseline point.
For example, requiring participants to complete all follow-up assessments before being considered part of the original study population can create selection problems.
Participants who withdraw or become unavailable may differ systematically from those who remain.
Similarly, restricting an intervention analysis to participants who adhered throughout the study may change the intended treatment-effect estimand.
Researchers should distinguish baseline eligibility from later retention, adherence, and missing outcome processes.
These later events require appropriate analytical handling rather than retrospective redefinition of initial eligibility.
How Do Inclusion Criteria Affect the Sampling Frame?
The sampling frame should identify units satisfying, or potentially satisfying, the eligibility criteria.
For example, a study targeting current undergraduate students may use an official enrollment register.
However, the register may include ineligible records or omit eligible students.
Researchers should assess whether the frame provides sufficient information to apply eligibility rules accurately.
When screening is necessary, the protocol should distinguish frame inclusion from confirmed study eligibility.
A suitable sampling method cannot correct a frame that systematically excludes eligible population members.
Should Inclusion Criteria Differ Across Study Groups?
In comparative studies, eligibility criteria should support a meaningful comparison between the groups.
For randomized trials, applying consistent baseline eligibility criteria before assignment helps preserve comparability.
For observational comparisons, different eligibility rules may create systematic differences that complicate interpretation.
However, some research questions legitimately concern groups defined by different characteristics or exposure histories.
The researcher should specify whether the criteria define comparable source populations and whether the intended contrast is meaningful.
Differences should be justified by the design rather than introduced merely to facilitate recruitment.
How Should Inclusion Criteria Be Written for Qualitative Research?
Qualitative inclusion criteria should identify participants whose experiences, roles, or circumstances are relevant to the research question.
For example, a study examining how university administrators implement AI assessment policies may recruit administrators who have participated in policy development or implementation.
Such criteria can support purposive sampling and information-rich cases.
However, eligibility should not be so restrictive that it excludes relevant contrasting experiences.
For example, including only administrators who support the policy could prevent the study from examining resistance or implementation difficulties.
Qualitative eligibility should reflect conceptual relevance and variation rather than statistical representativeness.
How Should Inclusion Criteria Be Written for Intervention Studies?
Intervention studies should identify participants for whom the intervention and intended outcome are meaningful and ethically appropriate.
For example, an AI tutoring intervention designed for introductory programming courses may require enrollment in those courses.
Additional eligibility conditions may concern the ability to receive the intervention, safety, baseline status, or other design requirements.
However, requiring participants to have prior success with the intervention or to demonstrate future adherence would change the population and potentially introduce selection bias.
Researchers should distinguish intervention eligibility from treatment uptake and adherence after assignment.
The criteria should align with the intended treatment-effect estimand.
How Do Inclusion Criteria Affect Generalizability?
Eligibility criteria determine which population members can contribute evidence to the study.
Restrictive criteria may produce a sample that differs from the population in which the findings will eventually be applied.
For example, excluding students with limited internet access from an AI tutoring evaluation may limit the relevance of findings to institutions where connectivity is less reliable.
However, broader eligibility does not automatically establish generalizability.
Researchers should consider whether excluded characteristics modify outcomes or treatment effects and whether the sampling design supports the intended population inference.
The implications of eligibility criteria for generalizability should be considered during protocol development.
Can Inclusion Criteria Be Too Restrictive?
Yes. Excessively restrictive criteria may reduce recruitment feasibility, remove relevant population variation, and limit the applicability of findings.
For example, a study of undergraduate AI literacy might require participants to be enrolled full-time, own a laptop, have reliable internet access, and possess previous AI experience.
Unless each restriction is justified, the resulting population may exclude students whose experiences are particularly relevant to digital learning inequalities.
Researchers should examine whether a criterion is necessary or whether the characteristic could instead be measured and analyzed.
A focused population is appropriate when the research question requires it, not merely because a narrow sample is easier to manage.
Can Inclusion Criteria Be Too Broad?
Yes. Insufficiently defined eligibility may include individuals for whom the phenomenon, intervention, or outcome is not meaningfully comparable.
For example, a study of undergraduate academic writing should not automatically include postgraduate students completing substantially different writing tasks.
However, heterogeneity is not inherently undesirable.
If the research aims to examine variation across educational levels, including those groups may be appropriate.
The question is whether the criteria preserve the intended estimand and support meaningful measurement and analysis.
How Should Borderline Eligibility Be Handled?
Borderline cases arise when individuals partially satisfy a criterion or when eligibility information is uncertain.
For example, a student may be transitioning between enrollment categories, or a participant's prior experience may not clearly satisfy an operational threshold.
Researchers should establish procedures for verification, clarification, and adjudication before recruitment.
Eligibility decisions should be documented and applied consistently.
However, automatically excluding every uncertain case may systematically remove participants with incomplete records or less conventional circumstances.
The protocol should distinguish genuine ineligibility from uncertainty in measuring eligibility.
Can Inclusion Criteria Be Changed During Recruitment?
Changes may sometimes be justified by new evidence, ethical concerns, or recruitment feasibility.
However, modifying eligibility criteria can change the study population and potentially the estimand.
For example, expanding an intervention study from second-year students to all undergraduate students may introduce different educational contexts and baseline characteristics.
Researchers should document the rationale, timing, and implications of any change.
For registered studies, relevant protocol and registration updates should follow applicable requirements.
Changes should not be introduced selectively after examining outcomes to obtain more favorable results.
How Can You Audit Inclusion Criteria Against the Research Question?
A useful audit asks whether each criterion is necessary for the population, intervention, measurement, or ethical requirements of the study.
For each criterion, identify what would happen if it were removed.
If removing the criterion would include individuals for whom the research question is still meaningful, the restriction may be unnecessary.
If removing it would make the outcome uninterpretable or the intervention inappropriate, the criterion may be justified.
Researchers should also examine whether the criterion excludes a subgroup central to the intended population.
This audit helps prevent eligibility rules from becoming a collection of convenient but scientifically disconnected restrictions.
How Should Inclusion Criteria Be Reported?
Methods sections should state inclusion criteria explicitly and describe how each was assessed.
Where relevant, identify the source of eligibility information, reference period, operational thresholds, and procedures for uncertain cases.
Explain important restrictions and their relationship to the target population.
For observational studies, STROBE emphasizes transparent reporting of eligibility criteria, participant selection, and study settings.
For randomized trials, CONSORT provides relevant guidance concerning eligibility and participant flow.
Clear reporting allows readers to determine whether the recruited population supports the research question and intended conclusions.
