03 · What You Need to Know
How Emerging Analysis Should Inform Qualitative Sampling Decisions
Why Might Qualitative Sampling Need to Change?
Qualitative researchers often begin with provisional understandings of a phenomenon. Initial sampling decisions reflect the research question, relevant literature, methodological approach, and available access.
As data collection and analysis proceed, the researcher may identify experiences or processes that were not anticipated.
For example, a study of AI-generated feedback may initially emphasize instructors' technical experiences. Early interviews may reveal that institutional accountability influences how instructors evaluate the technology.
The researcher may then need evidence from participants who encountered different accountability arrangements.
Such adaptation can strengthen the relationship between the sample and the developing analysis.
However, the need for change must be demonstrated rather than assumed simply because qualitative research is flexible.
Does Qualitative Research Always Require Adaptive Sampling?
No. Some qualitative designs use relatively stable eligibility criteria and recruitment procedures.
For example, a phenomenological study may focus on participants sharing a clearly defined lived experience. The original criteria may remain appropriate throughout data collection.
Other approaches, particularly grounded theory, place greater emphasis on iterative sampling informed by emerging analysis.
Case studies may also adapt data collection when additional actors or documents become relevant to understanding a bounded case.
The extent of flexibility should therefore be consistent with the methodological tradition.
Researchers should not assume that every qualitative study must revise its sampling plan to demonstrate rigor.
What Is the Difference Between Planned and Adaptive Sampling?
Planned sampling establishes initial criteria and recruitment procedures before data collection.
Adaptive sampling allows subsequent selection decisions to respond to information obtained during the research.
Initial sampling plan
Defines the starting population, eligibility criteria, relevant characteristics, and recruitment procedures.
Analytically adaptive sampling
Modifies or extends selection according to questions, gaps, or conceptual needs emerging from the evidence.
These approaches are not mutually exclusive. A study may begin with a carefully specified purposive sample and later introduce additional cases for clearly documented analytical reasons.
The initial plan provides coherence, while justified adaptation allows the inquiry to respond to what is learned.
When Do Emerging Categories Justify Additional Recruitment?
Additional recruitment may be justified when an emerging category remains insufficiently developed or its relationship with other categories is unclear.
Suppose early interviews suggest that perceived institutional trust influences instructors' willingness to disclose AI use.
The researcher may need participants who have experienced different levels of institutional support or policy uncertainty to examine that relationship.
The selection should be guided by a specific analytical question.
For example, the researcher might seek instructors who report high trust but low willingness to disclose AI use.
Such cases could clarify whether trust alone explains disclosure or whether additional conditions are involved.
How Is Adaptive Sampling Different from Theoretical Sampling?
Adaptive sampling is a broad description of recruitment that changes in response to developing research needs.
Theoretical sampling has a more specific methodological meaning, particularly in grounded theory.
It involves selecting additional evidence according to the needs of emerging theoretical categories and relationships.
For example, a grounded theory researcher may recruit cases specifically to develop the properties of an emerging category concerning professional accountability.
By contrast, a qualitative case study researcher may add technical support personnel after discovering that system maintenance is important to understanding institutional implementation.
Both decisions may be analytically justified, but the second is not automatically theoretical sampling.
Researchers should use terminology consistent with their methodological approach.
When Should Missing Participant Perspectives Trigger Change?
Recruitment changes may be necessary when the analysis reveals that essential perspectives are absent.
For example, a study of institutional AI policy implementation may initially interview administrators and instructors.
Emerging accounts may suggest that students interpret the policy differently from those responsible for implementing it.
If student interpretation is central to the research question, excluding students may leave an important gap.
However, the researcher should not automatically recruit every stakeholder group mentioned during interviews.
The process of identifying essential participant perspectives requires evaluating whether a group can contribute evidence necessary for the intended analysis.
When Should Contradictory Cases Trigger Additional Sampling?
Contradictory cases may reveal limitations in an emerging interpretation.
Suppose most instructors describe institutional guidance as increasing confidence in AI-supported assessment.
One instructor reports that detailed guidance increased anxiety because it emphasized possible sanctions for inappropriate use.
The researcher may need additional cases to examine whether policy clarity and perceived institutional risk operate differently under particular conditions.
However, additional recruitment is not always necessary. Existing interviews or documents may contain sufficient evidence to investigate the discrepancy.
Negative and deviant case sampling is most useful when it responds to an identifiable analytical problem.
Can Changes in Context Justify Sampling Adaptation?
Yes. Emerging analysis may reveal that the phenomenon operates differently across settings.
For example, instructors may describe substantially different AI assessment practices depending on whether institutional policies are permissive, restrictive, or ambiguous.
If the research question concerns how institutional conditions shape practice, additional settings may be necessary.
However, expanding the sample across many contexts can increase analytical complexity.
Researchers should evaluate whether the study can collect sufficient evidence within each added context.
Adaptation should not become uncontrolled expansion of the research scope.
When Should Sampling Become More Focused?
Adaptive sampling does not always mean recruiting more diverse participants.
Sometimes early analysis reveals that the original sample is too broad.
For example, a study of educators' AI experiences may include teachers, administrators, and technical personnel whose accounts concern substantially different processes.
The researcher may determine that the central analytical contribution concerns instructors' decisions about assessment feedback.
Subsequent recruitment could focus on instructors with direct experience of those decisions.
Such narrowing should be justified through the research question and analysis, not used to exclude inconvenient findings.
The issue of excessive heterogeneity in qualitative sampling is relevant when the original sample introduces more variation than the study can meaningfully examine.
When Should Sampling Become More Diverse?
Conversely, analysis may reveal that the sample is too narrowly concentrated in one experience or context.
For example, all initial participants may be instructors who successfully adopted AI-generated feedback.
If the research seeks to explain sustained adoption and discontinuation, the absence of unsuccessful implementation experiences becomes consequential.
Additional recruitment could include instructors who discontinued AI use or encountered substantial implementation barriers.
However, diversity should be introduced along dimensions relevant to the emerging analytical questions.
The goal is not to maximize participant differences indiscriminately.
How Can Analytical Memos Guide Sampling Changes?
Analytical memos provide a record of developing interpretations, uncertainties, and decisions.
They can help researchers explain why particular additional cases are needed.
A useful sampling memo may identify the emerging category, describe the evidence currently available, specify what remains unclear, and propose a type of case capable of addressing the gap.
For example, a memo might record that instructors' perceptions of institutional accountability appear related to AI disclosure decisions, but evidence from settings with conflicting policies remains limited.
The researcher could then justify seeking participants with experience of such policy conditions.
Memos also help distinguish analytical adaptation from retrospective rationalization.
What Should a Sampling Decision Record Contain?
Researchers can document changes through a structured sampling decision record.
| Decision element |
Question to document |
| Emerging finding |
What pattern, category, or discrepancy prompted reconsideration? |
| Analytical gap |
What remains insufficiently understood? |
| Proposed cases |
Which participants, settings, or documents could address the gap? |
| Selection rationale |
Why are these cases relevant to the developing analysis? |
| Methodological implications |
Does the change affect eligibility, scope, or the unit of analysis? |
| Resulting contribution |
How did the additional evidence affect the interpretation? |
The record need not be administratively elaborate, but it should preserve the reasoning behind consequential decisions.
Transparent documentation is particularly important when recruitment evolves over several analytical cycles.
Can Researchers Change Eligibility Criteria?
Eligibility criteria may sometimes require revision when emerging analysis reveals that the original definition does not adequately capture the phenomenon.
For example, a study initially restricted to instructors who permanently discontinued AI feedback may discover that temporary suspension is an important part of the discontinuation process.
The researcher might consider including instructors who suspended use temporarily.
However, changing eligibility criteria affects the boundaries of the sample and potentially the interpretation of earlier cases.
Researchers should document the reasons, apply revised criteria consistently where appropriate, and consider whether the change requires an ethics amendment.
Criteria should not be modified merely to admit convenient participants or remove cases that challenge preferred findings.
What Ethical Requirements Apply to Adaptive Sampling?
Adaptive recruitment remains subject to applicable research ethics and institutional requirements.
Changes involving new participant groups, sensitive information, recruitment channels, or data collection procedures may require prior approval or amendment.
For example, expanding a faculty interview study to include students may introduce different power relationships and consent considerations.
Researchers should not assume that an initial ethics approval automatically covers every subsequent sampling change.
The applicable requirements depend on the institution, jurisdiction, and approved protocol.
Methodological flexibility does not remove the obligation to protect participants and follow authorized procedures.
How Do You Avoid Changing Sampling to Confirm Preferred Findings?
Adaptive sampling can be misused when researchers selectively recruit participants who support an emerging explanation.
For example, a researcher convinced that AI feedback improves assessment efficiency might seek only additional successful adopters.
This could reinforce the initial interpretation without examining relevant contradictory experiences.
To reduce this risk, sampling decisions should follow explicit analytical questions and consider cases capable of challenging the current explanation.
Researchers should document alternative interpretations and avoid treating agreement as the sole indicator of analytical progress.
Reflexive examination of researcher assumptions and recruitment relationships remains important.
How Does Adaptive Sampling Affect Sample Size?
Adaptive sampling may increase, decrease, or redirect recruitment depending on emerging analytical needs.
There is no universal number of additional participants required after a sampling change.
Malterud, Siersma, and Guassora's information power framework emphasizes the relationship between sample adequacy, study aim, participant specificity, theoretical background, dialogue quality, and analytical strategy.
Researchers should evaluate whether additional cases contribute evidence necessary for the intended interpretation.
Increasing sample size without addressing the analytical gap may add little value.
Similarly, stopping recruitment because a predetermined number has been reached may be inappropriate when important conceptual questions remain unresolved.
When Should Adaptive Recruitment Stop?
Recruitment should stop according to the methodological approach and the adequacy of evidence for the intended analysis.
In grounded theory, theoretical saturation may inform decisions about whether emerging categories and relationships require further development.
Other qualitative traditions may use different adequacy criteria.
Researchers should distinguish the repetition of descriptive codes from sufficient understanding of meanings, conditions, and relationships.
Practical constraints may also limit further recruitment.
When the available evidence remains incomplete, researchers should acknowledge the limitation rather than claim that an adequacy criterion has necessarily been achieved.
Can Adaptive Sampling Undermine Comparability?
It can, particularly when recruitment changes alter the participant population or the meaning of the phenomenon being studied.
For example, adding administrators to a study originally focused on instructors' lived experiences may introduce accounts addressing a different unit of analysis.
The additional evidence may still be useful, but the researcher should explain how it relates to the original analytical framework.
Similarly, changing eligibility criteria midway through recruitment may create differences between earlier and later participants.
Researchers should consider whether the resulting cases remain analytically comparable or require separate treatment.
Adaptation should strengthen the inquiry rather than obscure the basis of comparison.
What If the Sampling Plan Was Preregistered or Approved in Advance?
Some qualitative studies have protocols, preregistrations, or institutional approvals specifying recruitment procedures.
Adaptive changes should be distinguished from the original plan and documented transparently.
Where preregistration is used, researchers should explain deviations and the analytical reasons for them.
Where ethics approval is required, relevant amendments should be obtained before implementing changes when mandated.
Adaptation does not necessarily undermine methodological integrity, but concealing deviations or presenting post hoc decisions as predetermined can do so.
The goal is to preserve a transparent record of how the study developed.
How Should Adaptive Sampling Be Reported in the Final Paper?
The methodology should describe the initial sampling strategy, recruitment criteria, and the role of concurrent analysis.
Researchers should identify important changes, the findings that prompted them, and how subsequent cases were selected.
The report should also explain whether the additional evidence refined, challenged, or expanded the interpretation.
Where recruitment changes affected the study's scope or comparability, those implications should be acknowledged.
Reporting frameworks such as COREQ and the Standards for Reporting Qualitative Research provide useful guidance for describing qualitative sampling and analytical procedures.
The strongest account makes the relationship between emerging analysis and sampling decisions visible rather than presenting the final sample as though it resulted from an entirely fixed plan.
