03 · What You Need to Know
What Determines the Number of Focus Groups a Study Requires?
Focus Group Sample Size Has Two Distinct Dimensions
Focus group studies involve decisions about both the number of participants within each group and the number of separate groups conducted.
These decisions are related, but they are not interchangeable.
A study involving four groups with six participants each generates different interactional conditions from one group containing 24 participants, even though the total number of participants is identical.
Separate groups provide opportunities to examine whether particular issues arise across discussions and how interpretations differ under different group conditions.
The decision about how many participants should be in a focus group concerns the interaction within a session. The decision about how many groups to conduct concerns the broader sampling and analytical coverage of the study.
Researchers should justify both decisions independently.
Why One Focus Group Is Often Insufficient
A single focus group can provide valuable qualitative evidence about how a particular set of participants discusses a topic.
However, its findings are closely tied to the composition, relationships, and interactional circumstances of that particular group.
For example, one discussion among teachers may produce apparent agreement about institutional AI policies because participants share similar experiences or because a particularly influential participant shapes the conversation.
Without additional groups, it may be difficult to examine whether the same interpretations emerge under different interactional conditions.
Multiple groups can help researchers investigate variation across discussions and assess whether findings depend heavily on one group's composition.
Nevertheless, a single focus group is not automatically invalid. It may be appropriate for a narrowly bounded exploratory activity or a study in which one specific group's interaction is itself the object of investigation.
The researcher should limit claims accordingly rather than treating one discussion as sufficient evidence about a broader population.
The Research Question Determines the Required Coverage
A narrowly focused question involving participants with relatively similar experiences may require fewer groups than a broad investigation involving multiple institutions, professional roles, or contrasting contexts.
For example, a study exploring how teachers in one department interpret a specific policy may have a more bounded scope than a study comparing AI governance experiences across several universities.
The second study may require groups representing different institutional arrangements because those differences are central to the research question.
However, the number of groups cannot be determined simply by counting the number of institutions.
Researchers must consider whether each context contains sufficient variation to warrant multiple discussions and whether the intended analysis involves explicit comparisons.
The key question is what range of group interactions and experiences must be examined to support the proposed interpretation.
Participant Homogeneity Can Affect the Number of Groups
When participants share relatively specific characteristics or experiences, discussions may reveal recurring issues within a smaller number of groups.
For example, a study involving teachers who completed the same institutional AI training program may have a more homogeneous participant population than one involving teachers from different educational sectors.
However, apparent similarity does not guarantee that participants interpret their experiences in the same way.
Researchers should consider whether relevant differences exist within the proposed population and whether the study needs to investigate them.
Decisions about homogeneous and heterogeneous focus group composition should therefore be made alongside the number of groups.
Homogeneity may facilitate shared discussion, but it does not automatically establish analytical adequacy.
Comparative Research May Require Multiple Groups per Category
When a study intends to compare participant categories, the sampling plan should provide sufficient evidence within each category.
For example, a researcher might compare how novice and experienced teachers interpret institutional AI policies.
Conducting one focus group with novice teachers and one with experienced teachers may reveal contrasting discussions, but it creates an interpretive difficulty.
Any observed difference could reflect the participant category, the particular people recruited, or the interactional dynamics of those individual groups.
Additional groups within each category may help researchers examine whether the apparent differences recur across discussions.
There is no universal minimum number of groups per category. The required coverage depends on the heterogeneity of participants, analytical objectives, and the strength of the comparisons intended.
Researchers should avoid making broad subgroup claims from a single discussion in each category without acknowledging these limitations.
What Does Empirical Research Say About the Number of Focus Groups?
Empirical methodological research provides useful evidence about how thematic information accumulates across focus groups.
Guest, Namey, and McKenna (2017) investigated the number of focus groups needed to identify themes in a particular qualitative dataset. Their findings suggested that a substantial proportion of commonly identified themes emerged within the first several groups.
However, the results were specific to the population, topic, analytical procedures, and operational definition of saturation used in the study.
They do not establish that every qualitative project requires the same number of discussions.
Studies with heterogeneous participants, multiple analytical categories, or complex interactional objectives may require different coverage.
Researchers should therefore use empirical findings as contextual evidence for planning rather than as numerical prescriptions.
Can Focus Group Sampling Continue Until Saturation?
In some qualitative approaches, researchers assess whether additional groups generate new relevant codes, thematic categories, or substantive understanding.
This can inform decisions about whether further recruitment is necessary.
However, saturation has different meanings across qualitative traditions, and it is not a universal requirement.
A study may identify recurring descriptive topics while still lacking sufficient evidence about disagreement, contextual variation, or the meanings participants attach to those topics.
For example, several groups may identify "unclear institutional guidance" as a concern, but additional discussions may reveal substantially different interpretations of what clarity requires.
Researchers should distinguish the recurrence of a code from adequate understanding of the issue.
The broader methodological question of whether qualitative sampling should continue until saturation is relevant, but focus group studies must additionally consider how group interaction shapes the evidence.
Why Saturation Should Not Be Treated as a Mechanical Stopping Rule
A statement such as "Focus groups continued until no new themes emerged" is insufficient unless the researcher explains what was assessed and how the decision was made.
Repeated topics may arise because the moderator asks similar questions or because the sample excludes participants with contrasting experiences.
Moreover, new interactions can generate different interpretations even when familiar topics recur.
Researchers should consider whether the analytical approach treats themes as discoverable categories or as interpretive constructions developed through engagement with the material.
In reflexive thematic analysis, for example, Braun and Clarke have questioned the conceptual compatibility of saturation with the approach's assumptions about theme development.
The stopping rationale should therefore be consistent with the methodology rather than adopted merely because saturation is commonly mentioned in qualitative research.
How Does Group Interaction Affect Sample Adequacy?
Focus groups generate data through participant interaction, so adequacy cannot be assessed solely by counting distinct opinions.
Researchers may need to examine whether discussions provide sufficient evidence about agreement, disagreement, negotiation, or collective interpretation.
For example, a group in which participants quickly agree may provide less evidence about contested policy interpretations than a discussion in which different positions are explored carefully.
However, disagreement should not be manufactured simply to make the discussion appear richer.
The researcher should examine how the interaction relates to the research question and whether additional groups would help investigate relevant differences.
Methodological adequacy concerns the interpretive contribution of the discussions, not merely the number of codes or statements generated.
Can Information Power Help Plan Focus Groups?
Information power, proposed by Malterud, Siersma, and Guassora (2016), provides a framework for evaluating qualitative sample adequacy in relation to the study aim, sample specificity, theory, dialogue quality, and analytical strategy.
Although developed in relation to qualitative interview studies, its underlying reasoning may inform focus group planning when applied cautiously.
For example, a narrowly focused study involving participants with highly relevant experiences may require fewer discussions than a broad study spanning several contexts.
However, focus group interaction introduces additional considerations that are not captured by simply transferring individual interview sample-size reasoning.
Researchers should therefore use information power as a conceptual aid rather than a formula for calculating the required number of groups.
How Should You Plan an Initial Number of Focus Groups?
A research proposal generally needs a credible estimate of the number of sessions, recruitment requirements, and analytical workload.
Researchers can establish a provisional target based on the scope of the research question, participant categories, expected variation, and available resources.
For example, a study comparing teachers from two institutional policy environments might initially plan several discussions within each environment.
The exact number should be justified by the intended comparisons rather than a universal numerical convention.
Where the methodology permits iterative sampling, the protocol may specify how analysis will inform additional recruitment.
Researchers should also establish practical boundaries, including the maximum recruitment permitted by ethics approval and the resources available for additional sessions.
Should All Focus Groups Have the Same Number of Participants?
Not necessarily.
Group size may vary because of recruitment, attendance, participant characteristics, or methodological considerations.
For example, one group may contain five participants and another seven.
Such variation does not automatically invalidate comparison, but it can influence the interactional conditions under which the discussions occur.
Researchers should consider whether differences in group size affected speaking opportunities, moderation, or the range of perspectives expressed.
Actual group sizes should be reported rather than assuming that every session matched its recruitment target.
How Does Focus Group Duration Affect the Number Needed?
The duration of each discussion influences how thoroughly participants can explore the intended topics.
A short session may identify initial reactions but provide limited opportunities to examine disagreement or contextual detail.
A longer session may support more extensive discussion, although fatigue and scheduling constraints must be considered.
However, longer groups do not necessarily replace the value of separate discussions involving different participants.
Multiple groups can provide evidence about variation across interactional settings, while extending one session primarily increases the time available within that setting.
Researchers should therefore plan session duration and group count as related but distinct design decisions.
What If One Group Produces Much Richer Discussion Than Another?
Differences in discussion quality are common.
Some groups may engage in sustained disagreement and detailed explanation, while others provide shorter or more cautious contributions.
Researchers should investigate whether these differences reflect participant composition, moderator practices, topic relevance, or social relationships.
A less interactive group is not automatically unusable. Its limited interaction may itself be informative, particularly when the research concerns professional norms or institutional hierarchy.
However, if the group failed to address essential topics because of procedural problems, an additional session may be necessary.
The decision should follow the analytical purpose rather than an assumption that every transcript must contain a similar amount of material.
Can Online and In-Person Groups Be Counted Together?
They can be included within the same study when the design justifies combining administration modes.
However, online and in-person discussions may differ in turn-taking, nonverbal communication, accessibility, and participant engagement.
These differences can affect the interactional evidence.
Researchers should consider whether online and in-person focus groups produce sufficiently comparable data for the intended analysis.
When modes differ, the methods section should describe how those differences were handled rather than treating every group as procedurally identical.
How Do Recruitment and Attendance Affect the Plan?
The number of sessions planned may differ from the number successfully conducted.
Participants may decline invitations, cancel, or fail to attend.
Researchers should prepare recruitment and confirmation procedures that reflect the population and ethical requirements.
However, over-recruitment should be managed carefully because inviting too many participants may create difficulties if everyone attends.
Researchers should establish procedures for rescheduling, arranging additional sessions, or adjusting group composition when attendance differs from expectations.
Recruitment feasibility is a legitimate design consideration, but difficulty recruiting participants does not itself demonstrate analytical adequacy.
When Should Additional Focus Groups Be Conducted?
Additional groups may be justified when important participant categories remain underrepresented, contextual comparisons are inadequately developed, or emerging interpretations require further examination.
For example, a study may discover that teachers from institutions with restrictive AI policies describe different experiences from those in more permissive environments.
If the contrast becomes central to the analysis, further groups may be needed to examine it adequately.
However, researchers should avoid expanding recruitment indefinitely whenever a new topic appears.
The relevance of additional evidence should be evaluated against the research question, analytical approach, ethical approval, and available resources.
How Should the Number of Focus Groups Be Reported?
The methods section should report the number of groups planned and conducted, the actual number of participants in each group, and the sampling rationale.
Researchers should explain how groups were composed and whether particular participant categories or contexts were examined separately.
If the number changed during fieldwork, describe why.
Where saturation or another adequacy criterion informed the stopping decision, explain how it was evaluated.
Reporting frameworks such as COREQ encourage transparency about sampling, data collection, and the research context.
The objective is to make the relationship between the achieved discussions and the study's analytical claims understandable.
