Focus groups are a qualitative research method in which a moderator guides a small group discussion about a defined topic. Researchers use them to explore participants’ experiences, language, attitudes, disagreements, and shared meanings. Their distinctive value comes from interaction: participants respond not only to the moderator but also to one another.
Focus groups are used in social science, education, healthcare, public policy, market research, program evaluation, communication, and user-experience research. They can help researchers understand why people hold particular views, how opinions develop in a social setting, and where participants agree or disagree.
This guide explains what focus groups are, when they are appropriate, how to recruit participants and write questions, how to moderate and analyze a discussion, and how to address sampling, ethics, online research, reporting, digital tools, and artificial intelligence.
Key takeaways
- A focus group is more than several interviews conducted at the same time; interaction among participants is an important source of data.
- Focus groups are best for exploring perceptions, language, norms, shared experiences, disagreement, and reactions to ideas.
- There is no universal rule for the correct number of participants or groups.
- Findings are normally interpreted qualitatively and should not be treated as statistically representative percentages.
- Confidentiality can be requested but cannot be fully guaranteed because participants hear one another.
- Technology and AI may assist transcription and organization, but researchers remain responsible for privacy, accuracy, interpretation, and methodological transparency.
What Are Focus Groups?
A focus group is a planned and moderated discussion involving a small number of participants selected because they have experiences, characteristics, or knowledge relevant to a research question.
The moderator introduces topics through a semi-structured discussion guide. Participants are encouraged to speak to one another, compare experiences, explain disagreements, question assumptions, and develop or reconsider ideas during the conversation.
Focus groups usually generate qualitative data, including:
- Spoken accounts and explanations
- Stories and examples
- Agreements and disagreements
- Participants’ language and terminology
- Reactions to other participants
- Changes in position during the discussion
- Silence, hesitation, humor, tension, and enthusiasm
- Relevant nonverbal behavior, when it can be observed ethically and reliably
Kitzinger (1995) emphasized that participant interaction is a defining methodological strength of focus groups. The researcher can examine not only what people think but also how opinions are expressed, supported, challenged, or negotiated in a social setting.
What Makes a Focus Group Different From a Group Interview?
The terms focus group, focus group interview, and group interview are sometimes used interchangeably. However, a methodologically strong focus group does not simply ask each participant the same question in turn.
In an ordinary group interview, the main communication may occur between the interviewer and each individual. In a focus group, the moderator deliberately encourages communication among participants.
For example, instead of asking five students separately whether online learning was effective, the moderator might ask:
“What was online learning like for students in your program?”
After one student responds, the moderator can invite interaction:
“Did others have a similar experience?”
“What made your experience different?”
“You mentioned flexibility. Did flexibility always improve learning?”
“Has anyone changed their view after hearing these examples?”
The interaction reveals variation, assumptions, group norms, and the conditions under which experiences differ.
What Is the Purpose of a Focus Group?
The purpose of a focus group is to develop an in-depth understanding of how a relevant group perceives, discusses, and makes sense of a defined topic.
Focus groups are particularly useful for:
- Exploring a topic about which little is known
- Identifying participants’ vocabulary and concepts
- Understanding attitudes, needs, expectations, and experiences
- Examining agreement, disagreement, and social norms
- Developing survey questions or intervention materials
- Evaluating a program, service, course, policy, product, or message
- Testing reactions to several concepts or prototypes
- Interpreting findings from a previous survey
- Generating hypotheses or propositions for later research
- Involving communities or service users in research and design
A focus group is not normally intended to produce a vote, force consensus, or estimate the percentage of an entire population holding an opinion.
When Should Focus Groups Be Used?
Use focus groups when the research question concerns shared language, collective experiences, social expectations, reactions, or the ways people influence one another.
They are suitable when participants can reasonably discuss the topic together and when hearing other perspectives is likely to stimulate useful reflection.
Appropriate focus group questions
Examples include:
- How do first-year students describe the challenges of adapting to university?
- How do nurses discuss barriers to implementing a new hospital procedure?
- What language do parents use when evaluating school communication?
- How do customers compare alternative versions of a product package?
- How do remote employees negotiate expectations about availability?
- How do community members differ in their understanding of a public-health message?
When focus groups may not be appropriate
A focus group may be unsuitable when:
- The research requires private, highly personal accounts.
- Disclosure could create legal, social, professional, or physical risk.
- Participants have strong power differences that cannot be managed through segmentation.
- The topic is likely to produce hostility rather than productive discussion.
- The researcher needs statistically generalizable prevalence estimates.
- Responses must be independent of other participants.
- The task concerns actual behavior that would be better studied through observation or usability testing.
- Participants cannot communicate safely or comfortably in the same setting.
- The researcher lacks the time or resources to recruit, moderate, transcribe, and analyze group interaction properly.
Sensitive topics are not automatically unsuitable. In some circumstances, participants feel supported by others with similar experiences. However, researchers must consider confidentiality, distress, stigma, safeguarding, and whether an individual interview would reduce risk.
Focus Groups vs. Interviews, Surveys and Workshops
| Method | Best used for | Main form of data | Interaction among participants | Main limitation |
|---|---|---|---|---|
| Focus group | Exploring shared meanings, norms, language, disagreement, and reactions | Group discussion and interaction | Central to the method | Social influence may suppress or reshape individual views |
| Individual interview | Exploring personal experiences, motivations, sensitive subjects, and detailed life accounts | One-to-one conversation | None | Does not directly show how views are negotiated socially |
| Survey | Measuring distributions, frequencies, associations, or standardized responses | Numerical or structured response data | None | Limited depth and flexibility |
| Workshop | Producing ideas, decisions, plans, designs, or collective outputs | Discussion plus task outputs | Usually central | The goal is often action or co-creation rather than research data alone |
| Observation | Studying behavior, routines, settings, and interaction as they occur | Field notes, recordings, or behavioral records | Naturally occurring | Internal reasoning may remain unknown |
Focus groups and interviews are complementary rather than competing methods. A project may begin with focus groups to identify shared concerns and then use interviews to explore private experiences. Alternatively, interviews may identify themes that are later discussed collectively.
Types of Focus Groups
Traditional in-person focus groups
Participants meet in the same physical setting. This format supports natural turn-taking, observation of body language, use of physical materials, and informal conversation before and after the formal session.
Online synchronous focus groups
Participants meet in real time using secure video, audio, or text-based technology. These groups can include geographically dispersed participants but require careful planning for privacy, connectivity, accessibility, screen size, and turn-taking.
Asynchronous online focus groups
Participants respond over a longer period through a protected discussion board or similar platform. This gives participants time to reflect and can reduce scheduling problems, but it changes the pace and spontaneity of interaction.
Mini focus groups
Mini groups usually contain fewer participants than conventional groups. They are useful when:
- The topic requires additional speaking time.
- Participants have specialist knowledge.
- Recruitment is difficult.
- The subject is sensitive.
- Online turn-taking would become difficult in a larger group.
Dual-moderator focus groups
One moderator manages the discussion while another ensures that the research topics are covered. Roles must be clearly defined so participants do not receive conflicting direction.
Two-way focus groups
One group observes or reviews material from another group and discusses its reactions. This format may be useful for studying interpretations of group communication but requires a clear rationale and additional consent.
Expert focus groups
Participants are selected because of professional, technical, policy, or experiential expertise. Researchers must avoid assuming that experts share one unified perspective.
Repeated or longitudinal focus groups
The same participants meet more than once. Repeated groups can examine changing experiences, implementation processes, or responses over time. However, familiarity may alter group dynamics and disclosure.
Deliberative or participatory groups
Participants receive information, examine alternatives, or help develop recommendations. These groups can overlap with participatory research and workshops. Researchers must clarify whether the purpose is data collection, co-design, decision-making, or a combination.
How to Conduct a Focus Group
A defensible focus group study requires more than inviting several people to answer questions. The study should connect its purpose, sample, discussion guide, moderation, analysis, and claims.
Step 1: Define the Research Purpose
Begin with a focused statement explaining what the study needs to understand and why group interaction is useful.
A strong purpose statement might be:
“This study will explore how postgraduate students discuss the benefits, barriers, and social expectations associated with using generative AI for academic writing.”
The purpose should guide:
- Who should participate
- Whether separate groups are required
- Which topics should be discussed
- What materials should be presented
- Which analytical approach will be used
- What conclusions the study can reasonably support
Avoid choosing focus groups merely because they appear faster than interviews. Recruitment, scheduling, moderation, transcription, interaction-sensitive analysis, and reporting may require substantial work.
Step 2: Develop Research Questions
Research questions state what the project seeks to understand. They are not necessarily the exact questions that the moderator will read aloud.
Example research questions:
- How do postgraduate students understand acceptable and unacceptable uses of generative AI?
- Which institutional expectations influence their decisions?
- Where do students agree or disagree about authorship and academic integrity?
- How do disciplinary experiences shape their views?
A moderator question should be more conversational:
“When students in your program talk about acceptable uses of AI, what examples usually come up?”
Step 3: Select Participants
Focus groups usually use purposive sampling. Participants are selected because they can contribute relevant experiences or perspectives.
Possible inclusion criteria include:
- Experience using a service or program
- Membership of a professional or community group
- A particular educational stage
- Exposure to a policy or intervention
- Relevant demographic or contextual characteristics
- Specialist or lived experience
Researchers should clearly state:
- Inclusion and exclusion criteria
- Recruitment channels
- Who invited participants
- Whether incentives were offered
- Reasons for nonparticipation, when known and relevant
- Whether participants knew one another
- The relationship between participants and the research team
Homogeneous or heterogeneous groups?
A homogeneous group shares characteristics that may support comfort and meaningful comparison. For example, junior nurses might be placed in a group separate from senior managers.
A heterogeneous group brings together varied perspectives. This can reveal disagreement but may also create status differences or inhibit disclosure.
Homogeneity should be relevant to the research question rather than interpreted as making participants identical. A group can be similar in professional status while varying in age, experience, gender, location, or opinion.
When to segment groups
Consider separate groups when combining participants might discourage honest discussion. Examples include:
- Employees and supervisors
- Students and instructors
- Patients and healthcare professionals
- Parents and adolescents
- Service users and service providers
- Junior and senior professionals
- People with sharply different levels of authority
Segmentation may improve psychological safety, but it also changes the comparisons that can be made. The rationale should be documented.
Step 4: Determine Group Size
There is no universally correct group size.
Many in-person focus groups contain approximately six to ten participants, although smaller or larger groups may be justified. The appropriate number depends on:
- Complexity of the topic
- Desired speaking time
- Participant characteristics
- Sensitivity of the discussion
- Moderator experience
- Use of visual or task-based activities
- Online or in-person format
- Accessibility and communication needs
- Expected attendance and no-shows
| Situation | Practical group-size consideration |
|---|---|
| Complex or sensitive topic | A smaller group may allow deeper and safer discussion |
| Highly experienced participants | A smaller group can provide more time for specialist accounts |
| Broad exploratory discussion | A moderate-sized group may generate varied reactions |
| Online video session | Fewer participants may improve visibility and turn-taking |
| Children or participants needing support | Smaller groups and additional safeguards may be appropriate |
| Short concept-testing task | A somewhat larger group may work if individual depth is not the main goal |
A planned group of ten does not guarantee ten attendees. Recruitment plans should account for likely absences without deliberately overcrowding the session.
How many focus groups are needed?
The number of groups is different from the number of people in each group.
Researchers should consider:
- The number of relevant subgroups
- Whether comparisons among sites or populations are required
- The specificity of the research questions
- The diversity of experiences
- The type and depth of analysis
- Whether data collection and analysis are iterative
- Available time and resources
- The intended claims
Some empirical studies suggest that four to eight focus group discussions can be sufficient to reach forms of saturation in relatively homogeneous samples with narrow objectives. This should not be treated as a universal formula (Hennink & Kaiser, 2022).
A study seeking code saturation may require fewer groups than one seeking a rich understanding of the meanings and variations within each theme. Researchers should explain their stopping rationale instead of stating only that “saturation was reached.”
One focus group can provide useful pilot, consultation, or local feedback data, but it usually offers limited evidence about whether the patterns are specific to that session’s composition and dynamics.
Step 5: Address Ethics and Consent
Ethical planning should begin before recruitment.
Participants should receive understandable information about:
- The study’s purpose
- What participation involves
- Session duration
- Topics likely to be discussed
- Audio or video recording
- Data storage and access
- Use of quotations
- Voluntary participation
- The right to decline questions
- Withdrawal procedures
- Incentives or reimbursements
- Potential benefits and risks
- Research-team contact information
- Complaints or ethics-review procedures
Confidentiality in a group setting
The research team can protect data under its control, but it cannot guarantee that participants will keep everything confidential.
The consent process should distinguish between:
- Researcher confidentiality: how the team stores, anonymizes, and reports information.
- Participant confidentiality: the request that participants do not repeat what others say.
Ground rules should include a clear request not to share other participants’ identities or comments outside the session. Researchers should still avoid encouraging disclosures that could create serious risk.
Recording and withdrawal
Researchers should state whether audio, video, chat messages, screen activity, or observation notes will be collected.
Withdrawal becomes complicated after group data have been combined, anonymized, coded, or quoted. A participant may be able to leave the session immediately, but removing every influence of that person’s comments from an interactive discussion may be impossible. This limitation should be explained before consent.
Step 6: Create the Discussion Guide
A discussion guide organizes topics without turning the session into a rigid questionnaire.
A common sequence is:
- Welcome and introduction
- Purpose and ground rules
- Easy opening question
- Introductory experience questions
- Transition questions
- Key questions
- Follow-up probes and activities
- Summary and participant confirmation
- Final comments
- Closing and debriefing
Sample 75-minute session plan
| Time | Activity |
|---|---|
| 0–10 minutes | Welcome, consent confirmation, recording permission, introductions, and ground rules |
| 10–20 minutes | Opening questions and general experiences |
| 20–55 minutes | Key questions, probes, comparisons, and stimulus materials |
| 55–65 minutes | Prioritization, contrasting views, or unresolved issues |
| 65–72 minutes | Moderator summary and participant correction |
| 72–75 minutes | Final comments, support information, incentives, and closing |
Characteristics of effective questions
Effective questions are usually:
- Open-ended
- Neutral
- Clear
- Conversational
- Focused on one main idea
- Relevant to participants’ experience
- Capable of producing interaction
- Ordered from easier to more demanding topics
Avoid:
- Leading questions
- Double-barrelled questions
- Academic jargon
- Questions that presume an experience
- Questions that pressure consensus
- Long hypothetical scenarios
- Questions asking participants to predict behavior without evidence
- Excessive yes-or-no questions
- Questions likely to shame or expose participants
Sample questions and probes
Opening question
“Please tell us briefly about your experience with the university’s online-learning system.”
Experience question
“Think about the last time you used it for an important assignment. What happened?”
Comparison question
“How was that different from completing the same kind of task in person?”
Group-interaction question
“Several people have described convenience. Does anyone see convenience differently?”
Clarification probe
“What does ‘easy to use’ mean in practice?”
Example probe
“Can you describe a specific situation?”
Contrast probe
“Was there a time when the opposite happened?”
Minority-view invitation
“We have heard several similar views. What might someone who disagrees say?”
Closing question
“What important issue have we not discussed?”
The moderator does not need to use every probe. Probes should respond to the actual conversation.
Step 7: Prepare the Setting and Research Team
A typical team includes:
- Moderator: guides the discussion and manages interaction.
- Assistant moderator or note-taker: manages logistics, records seating or speaker information, observes dynamics, monitors equipment, and supports debriefing.
- Additional support staff: used only when required for accessibility, safeguarding, language interpretation, or technical support.
The setting should be:
- Private
- Accessible
- Neutral where possible
- Comfortable
- Quiet
- Safe
- Convenient
- Suitable for recording
- Arranged so participants can see and hear one another
Researchers should test equipment, prepare backups, label materials, confirm accessibility, and decide how unexpected observers or support persons will be handled.
Step 8: Moderate the Discussion
The moderator’s task is not to present personal expertise or persuade participants. It is to create a respectful setting in which relevant interaction can occur.
An effective moderator should:
- Explain the purpose without suggesting a preferred answer.
- Establish respectful ground rules.
- Invite participants to respond to one another.
- Ask neutral follow-up questions.
- Notice who has and has not spoken.
- Protect minority viewpoints.
- Manage time without rushing useful discussion.
- Separate disagreement from personal attack.
- Monitor distress or safeguarding concerns.
- Avoid completing participants’ sentences.
- Avoid praising one answer more than another.
- Summarize carefully and invite correction.
Managing a dominant participant
The moderator can:
- Thank the person and invite other views.
- Use a round-robin response for a specific question.
- Direct a question toward participants who have spoken less.
- Say, “Let us hear from someone who has not commented yet.”
- Speak privately during a break if domination continues.
The moderator should not embarrass or punish the participant.
Encouraging quieter participants
Silence does not necessarily indicate lack of knowledge. A participant may be reflecting, disagreeing, uncomfortable, or responding to status differences.
Helpful techniques include:
- Allowing pauses
- Asking for examples
- Inviting rather than demanding a response
- Using brief written reflection before discussion
- Using pairs before whole-group discussion
- Offering chat or anonymous response options online
- Addressing accessibility or language needs
Managing conflict
Disagreement can produce valuable data. The goal is not to eliminate it but to keep it relevant and respectful.
The moderator can say:
“We have two different interpretations. Let us examine what experiences lead to each one.”
Personal attacks, discriminatory comments, threats, and unsafe disclosures require stronger intervention according to the study’s safeguarding plan.
Step 9: Record and Document the Session
Audio recording is common, while video may be justified when nonverbal interaction is analytically important. Video collects more identifiable information and therefore requires additional ethical justification and security.
Useful documentation includes:
- Recording
- Speaker-labeled transcript
- Seating map for in-person groups
- Attendance information
- Field notes
- Notable nonverbal behavior
- Technical problems
- Moderator reflections
- Assistant-moderator notes
- Immediate debrief summary
- Decisions about changes to later sessions
Researchers should preserve the distinction between what was recorded, what was observed, and what was interpreted after the session.
How to Analyze Focus Group Data
Focus group analysis should answer the research questions while preserving the context of participant interaction.
1. Organize the data
Create a clear identifier for every group, participant, recording, transcript, note, and analytical memo. Replace direct identifiers with pseudonyms or participant codes.
2. Verify the transcript
Automated transcripts should be checked against the recording. Common errors involve:
- Overlapping speech
- Accents and dialects
- Technical vocabulary
- Similar voices
- Names and places
- Quiet speech
- Code-switching
- Emotional expression
- Background noise
The required level of transcription depends on the analysis. Conversation analysis requires more detail than a broad thematic summary.
3. Become familiar with the data
Read transcripts and notes in full before fragmenting them into codes. Note:
- Initial impressions
- Recurring ideas
- Unexpected issues
- Strong disagreements
- Apparent consensus
- Changes in energy or tone
- The influence of the moderator
- Differences among groups
4. Choose an analytical approach
Common options include:
- Thematic analysis
- Qualitative content analysis
- Framework analysis
- Grounded theory procedures
- Discourse analysis
- Conversation analysis
- Matrix-based applied analysis
The method should fit the research question and epistemological position. Thematic analysis is flexible but still requires researchers to explain how themes were developed and what counted as a meaningful pattern (Braun & Clarke, 2006).
5. Code the material
Codes identify features relevant to the research question.
Example excerpt:
“I liked working from home, but everyone expected an immediate reply because they could see I was online.”
Possible codes:
- Flexibility
- Digital visibility
- Immediate-response expectations
- Boundary erosion
- Availability pressure
Codes should not be created solely by counting repeated words. A rare comment may be analytically important.
6. Analyze interaction
Ask:
- Who introduced the issue?
- Did others agree, challenge, qualify, or ignore it?
- Did participants give examples?
- Did someone change position?
- Was agreement immediate or hesitant?
- Did laughter indicate comfort, embarrassment, resistance, or shared understanding?
- Did a dominant participant shape the discussion?
- Were minority views expressed only after direct invitation?
- Did different groups discuss the issue in different ways?
7. Compare within and between groups
A useful matrix can include:
- Themes in rows
- Focus groups in columns
- Summary of each group’s position
- Important quotations
- Disagreements
- Contextual notes
- Negative or deviant cases
- Researcher interpretations
Comparisons should not erase internal disagreement. “Group 2 supported the policy” may be misleading if two participants strongly disagreed and others remained silent.
8. Develop and review themes
Themes should explain patterned meaning relevant to the research question. They are not simply headings copied from the discussion guide.
For example:
Topic heading: Remote-work communication
Descriptive category: Messaging expectations
Analytical theme: Flexibility became a new form of constant availability
Researchers should review whether each theme:
- Has sufficient supporting data
- Has a clear central concept
- Is distinct from other themes
- Accounts for variation
- Addresses the research question
- Reflects both content and interaction where relevant
9. Maintain an audit trail
Document:
- Coding decisions
- Changes to the codebook
- Analytical memos
- Alternative interpretations
- Team discussions
- Reflexive observations
- Reasons for combining or separating themes
- How contradictory cases were treated
- How software or AI tools were used
10. Report evidence transparently
Use quotations to illustrate the analysis, not to replace it.
A strong results section explains:
- The analytical claim
- The context in which it occurred
- Supporting and contrasting evidence
- Relevant group interaction
- The researcher’s interpretation
- Connections to the research question and literature
Avoid reporting statements such as “70% of participants agreed” unless the design explicitly incorporated an appropriate quantitative component. Focus group participants are not normally a probability sample, and their responses are influenced by the group context.
Worked Academic Example
Research topic
Postgraduate students’ experiences of AI-assisted academic writing.
Research purpose
To explore how students define acceptable assistance, how university rules influence practice, and where disciplinary differences create uncertainty.
Sampling plan
The researcher organizes six groups:
- Two groups of humanities students
- Two groups of social-science students
- Two groups of computing students
Students and instructors are not placed in the same groups because assessment authority may restrict disclosure.
Discussion topics
- Current uses of AI tools
- Perceived benefits
- Risks and errors
- Authorship
- Disclosure
- University guidance
- Disciplinary differences
- Fairness
- Suggestions for support
Possible findings
The analysis identifies the theme “Rules without shared definitions.” Students repeatedly refer to following university policy, but they disagree about whether outlining, language correction, paraphrasing, and feedback constitute authorship assistance.
A second theme, “Efficiency accompanied by verification work,” shows that students value speed but describe substantial effort checking references, technical claims, and invented information.
The report should explain both the themes and the interaction through which participants challenged one another’s definitions.
Advantages of Focus Groups
Rich interaction
Participants can clarify, contest, extend, and reinterpret one another’s accounts.
Efficient exposure to varied perspectives
A researcher can hear several perspectives during one session, although the total project is not necessarily quicker or cheaper than individual interviews.
Participant-generated language
Focus groups reveal the terms, metaphors, categories, and examples participants use naturally.
Identification of norms and disagreement
Researchers can examine which views appear acceptable, contested, embarrassing, or taken for granted.
Flexibility
Moderators can ask follow-up questions and explore unexpected issues.
Useful for developing other methods
Findings can inform survey items, interview guides, interventions, educational materials, messages, and product concepts.
Immediate comparison
Participants can respond directly to alternative ideas, prototypes, images, policies, or messages.
Limitations of Focus Groups
Social desirability and conformity
Participants may express views they consider acceptable rather than their private opinions.
Dominant voices
Highly confident or high-status participants can shape the direction and apparent consensus.
Limited privacy
Participants hear one another’s contributions, so confidentiality cannot be guaranteed.
Unequal speaking time
Some participants may speak extensively while others contribute little.
Complex analysis
Researchers must interpret both the content and the interaction, including crosstalk and contradictory statements.
Moderator influence
Question wording, tone, sequencing, responses, and nonverbal behavior can shape the discussion.
Recruitment and scheduling
Bringing relevant participants together can be difficult, especially across locations, shifts, caring responsibilities, or accessibility needs.
Limited statistical generalization
Purposively selected participants cannot normally provide population estimates.
Group composition effects
A different combination of participants might produce a different conversation.
Unsuitability for some sensitive topics
Fear, stigma, professional risk, or trauma may prevent disclosure or create harm.
Common Focus Group Mistakes
Treating a focus group as a cheap survey
Focus groups explain perspectives and meanings. They do not provide reliable percentages for a population.
Asking every participant every question in turn
This creates a serial group interview and can suppress natural interaction.
Recruiting people with unresolved power differences
Employees may not criticize a policy in front of supervisors, and students may not discuss teaching honestly in front of instructors.
Conducting only one group without qualification
A single group may be useful, but researchers should acknowledge that its dynamics cannot be compared with other sessions.
Writing too many questions
A long guide encourages shallow answers. Prioritize a small number of key questions and allow time for interaction.
Asking leading or double-barrelled questions
Questions should not contain the preferred answer or ask about several issues at once.
Ignoring quiet or minority views
Apparent agreement may reflect silence, status, fatigue, or avoidance.
Relying on the transcript alone
Transcripts may omit body language, seating, emotional tone, interruptions, technical problems, and contextual observations.
Counting comments as independent votes
Repeated statements within one interactive group are not equivalent to independent survey responses.
Claiming saturation without evidence
Researchers should state what type of saturation or sufficiency they sought, how it was assessed, and how sampling strata affected the decision.
Allowing software to determine the interpretation
Software organizes data. It does not decide which meanings are theoretically or contextually important.
Online and Hybrid Focus Groups
Online focus groups may improve access for geographically dispersed participants, people with mobility constraints, or those balancing work and caring responsibilities.
However, they can also exclude participants with:
- Limited internet access
- Insufficient devices
- Low digital confidence
- Inaccessible software
- No private place to participate
- Hearing or visual accessibility needs
- Limited data allowance
- Unstable electricity or connectivity
Online preparation checklist
Before the session:
- Use an institutionally approved platform.
- Review the platform’s privacy and recording settings.
- Send clear access instructions.
- Offer a short technical test.
- Use passwords and waiting rooms.
- Clarify display-name requirements.
- Encourage headphones where appropriate.
- Explain whether participants may use virtual backgrounds.
- Prohibit unauthorized recording.
- Prepare a telephone or reconnecting procedure.
- Provide an accessibility contact.
- Decide how chat messages and reaction icons will be treated as data.
Online video groups often benefit from fewer participants because turn-taking and visual attention become harder as the number of screens increases (Lobe et al., 2020).
Hybrid groups, where some participants attend physically and others remotely, can create unequal participation. Remote participants may struggle to enter the conversation or interpret room dynamics. Use hybrid formats only when the team can provide suitable microphones, cameras, facilitation, and equal opportunities to contribute.
Digital Tools and Artificial Intelligence
Digital tools can support:
- Recruitment tracking
- Consent administration
- Secure recording
- Transcription
- Speaker labeling
- Data organization
- Coding
- Memo writing
- Search and retrieval
- Comparison across groups
- Data visualization
- Team collaboration
- Reporting audits
Qualitative data-analysis applications include NVivo, MAXQDA, ATLAS.ti, Dedoose, and other approved platforms. The best choice depends on institutional access, security, team workflow, analytical method, and researcher competence.
Can AI transcribe focus groups?
AI transcription may be useful, but focus group audio is particularly difficult because of:
- Overlapping speech
- Multiple similar voices
- Accents
- Rapid turn-taking
- Background noise
- Informal language
- Specialist terminology
- Code-switching
Every transcript used for analysis should be checked against the recording to the degree required by the method.
Can AI code or analyze focus group data?
AI can assist with organizational tasks such as:
- Suggesting provisional labels
- Retrieving passages
- Summarizing coded extracts
- Comparing draft code definitions
- Identifying material for human review
- Formatting matrices
It should not be treated as an autonomous qualitative researcher. Generative systems can:
- Miss context and irony
- Invent explanations
- Overemphasize common language
- Suppress minority interpretations
- Produce unstable results
- Reproduce cultural and linguistic bias
- Ignore nonverbal and relational context
- Misrepresent uncertainty
- Expose confidential data
Researchers remain responsible for the interpretation.
Responsible AI checklist
Before using AI with research data:
- Check ethics approval, consent, institutional policy, and data-protection requirements.
- Determine whether the tool stores prompts or uses data for model training.
- Do not upload identifiable transcripts or recordings to unauthorized public systems.
- Use secure, institutionally approved environments.
- Minimize and anonymize data where possible.
- Verify every transcript, quotation, code, and summary.
- Preserve human engagement with the complete data.
- Document the tool, version, purpose, prompts, checks, and researcher decisions.
- Examine whether AI use changes the analytical approach.
- Report material AI assistance transparently.
Synthetic AI-generated participants should not be presented as substitutes for real people’s experiences. Simulated responses may be useful for testing a draft question guide, but they do not constitute evidence about a human population.
How to Improve Rigor and Trustworthiness
A high-quality focus group study should demonstrate alignment among its research question, group composition, facilitation, analysis, and claims.
Useful strategies include:
Credibility
- Recruit participants with relevant experience.
- Encourage contrasting and minority views.
- Use participant examples rather than abstract agreement alone.
- Compare interpretations with complete transcripts and notes.
- Search for negative or deviant cases.
- Summarize major points during the session and invite correction.
Dependability
- Maintain a documented procedure.
- Record changes to the guide and sampling plan.
- Preserve coding decisions and analytical memos.
- Explain how multiple researchers contributed.
Confirmability
- Use reflexive notes.
- Distinguish participant accounts from researcher interpretation.
- Examine moderator influence.
- Retain an audit trail.
Transferability
- Describe the setting, participants, recruitment, group composition, and context sufficiently for readers to judge relevance elsewhere.
Reflexivity
Researchers should consider how their identity, institutional role, assumptions, language, status, and relationship with participants may have influenced:
- Recruitment
- Disclosure
- Questioning
- Group dynamics
- Coding
- Interpretation
- Quotation selection
- Reporting
How to Report a Focus Group Study
The methodology section should normally report:
- Research purpose and questions
- Methodological orientation
- Why focus groups were appropriate
- Research-team roles and relevant characteristics
- Relationship with participants
- Sampling strategy
- Inclusion and exclusion criteria
- Recruitment process
- Number of groups
- Participants per group
- Group segmentation
- Setting and format
- Session length
- Discussion-guide development
- Pilot testing
- Recording and field notes
- Ethical approval and consent
- Transcription procedures
- Analytical method
- Coding process
- Researcher reflexivity
- Use of software or AI
- Procedures for rigor
- Stopping rationale
- Participant quotations
- Contradictory findings
- Limitations
COREQ is a 32-item reporting checklist developed for interviews and focus groups (Tong et al., 2007). SRQR provides 21 broader standards for qualitative research reporting (O’Brien et al., 2014). These tools improve transparency, but they should not be treated as substitutes for sound study design.
Focus Group Quality Checklist
Before data collection, confirm that:
- The research question requires group interaction.
- Participant criteria are justified.
- Power differences have been considered.
- The planned group size fits the format and topic.
- The number of groups is linked to the sampling and analytical strategy.
- Consent materials explain the limits of confidentiality.
- Questions are open, neutral, and conversational.
- The guide contains enough time for participant interaction.
- The moderator and assistant know their roles.
- The setting and technology have been tested.
- Accessibility and safeguarding plans are in place.
- Recording and storage meet institutional requirements.
- An analysis plan exists before data collection.
- The reporting checklist has been reviewed.
- Any use of transcription software or AI has been approved and documented.
Conclusion
Focus groups are most valuable when researchers need to understand how people discuss, compare, challenge, and negotiate experiences or ideas. Their strength lies in interaction, not simply in collecting several answers simultaneously.
A rigorous study requires a justified sample, thoughtful group composition, neutral questions, skilled moderation, transparent ethics, interaction-sensitive analysis, and appropriately limited conclusions. Online platforms and AI tools can support the process, but neither removes the researcher’s responsibility for privacy, interpretation, reflexivity, and methodological quality.
