Can Generative AI Help Create a Focus Group Discussion Guide?

Can AI Help You Design a Focus Group Discussion That Encourages Meaningful Interaction?

You are planning a focus group discussion about university instructors' experiences with generative AI. You need a discussion guide that introduces the topic, encourages participants to share their perspectives, and helps the moderator explore similarities and differences within the group.

Generative AI can quickly produce a list of open-ended questions, organize them into sections, and suggest follow-up prompts. The resulting guide may resemble a semi-structured interview protocol.

However, a focus group is not simply an interview conducted with several people simultaneously. Its distinctive methodological value often comes from participants responding to one another, comparing experiences, questioning assumptions, and developing ideas through interaction.

Can AI help prepare a guide that supports these interactions rather than merely collecting a series of individual responses?

Yes, AI Can Draft a Focus Group Guide, but Group Interaction Must Shape Its Design

How Generative AI Can Support Focus Group Discussion Guide Development

What Makes a Focus Group Different From an Individual Interview?

A focus group is a qualitative data collection method in which a moderator facilitates discussion among participants about a defined topic or set of issues.

Although individual experiences remain important, focus groups also generate data through interaction. Participants may agree, disagree, elaborate on one another's accounts, or reconsider their perspectives during the discussion.

Kitzinger (1995) emphasizes the importance of interaction in focus group research, including how participants respond to one another and collectively discuss experiences and understandings.

For example, an instructor may describe using AI to prepare assessments. Another participant may respond that their department discourages the practice. The ensuing discussion could reveal how institutional expectations differ across departments or how participants interpret those differences.

Such interaction may provide insights that would not emerge in the same way through separate individual interviews.

The methods are not mutually exclusive, and both can contribute valuable qualitative data. The choice should reflect the research questions and the kind of evidence the study seeks.

Start With the Purpose of the Group Discussion

Before asking AI to generate a guide, researchers should establish why a focus group is appropriate for the study.

A focus group may be useful when the research seeks to understand shared experiences, contrasting perspectives, collective interpretations, or how people discuss a phenomenon in a social setting.

For example, a study might investigate how university instructors discuss institutional expectations concerning generative AI. The interaction among participants could reveal areas of agreement, contested interpretations, or differences in professional practice.

However, a focus group may be less suitable when participants are expected to disclose highly personal experiences, confidential information, or views that could expose them to professional consequences in front of colleagues.

AI can suggest a focus group structure, but it cannot determine whether group discussion is ethically and methodologically preferable to individual interviews in the actual research setting.

What Can AI Help Develop for a Focus Group?

AI can support several practical components of focus group preparation, provided the researcher supplies the study's purpose, participant characteristics, and relevant contextual constraints.

These outputs are preparation materials. They do not establish that the group will interact as anticipated or that the questions will produce useful qualitative data.

Do Not Simply Convert Interview Questions Into Group Questions

One limitation of AI-generated focus group guides is that they may resemble individual interview guides with plural pronouns.

For example, an AI system might change "Can you describe your experience using AI?" into "Can everyone describe their experience using AI?"

This may encourage participants to answer sequentially while interacting primarily with the moderator.

A more interaction-oriented question might be: "What similarities or differences have you noticed in how instructors in your departments approach generative AI?"

The moderator could then invite participants to respond to perspectives introduced by others.

However, interaction-oriented questions should not assume that participants know one another's practices or are comfortable discussing identifiable colleagues.

The researcher should design questions that make group interaction possible without requiring participants to disclose information they cannot appropriately share.

Use Questions That Invite Comparison Without Forcing Consensus

Focus groups can help researchers explore how participants interpret similarities and differences in their experiences.

For example, after several instructors describe different approaches to AI-assisted assessment, the moderator might ask: "How do these experiences compare with what others in the group have encountered?"

This invites interaction without assuming that the group should reach agreement.

By contrast, asking "Can everyone agree on the main problem with AI assessment?" encourages consensus and presupposes that a shared problem exists.

Researchers should avoid treating consensus as the necessary outcome of a successful focus group. Disagreement, uncertainty, and contrasting interpretations may be equally informative.

AI can suggest comparison prompts, but the moderator must determine when they are appropriate during the actual discussion.

How Should You Prompt AI to Create a Focus Group Discussion Guide?

A useful prompt should identify the research questions, participant group, intended discussion duration, and methodological purpose of using a focus group.

This prompt encourages AI to consider group interaction as part of the research design.

The researcher should still examine whether the questions reflect the actual participant population and whether the discussion format is suitable for the topic.

What Types of Questions Work Well in Focus Groups?

Focus group questions may serve different functions during the discussion. Some introduce the topic, while others invite experiences, comparison, or reflection.

The examples should not be treated as a universal sequence. The appropriate questions depend on the research purpose and how the discussion develops.

What Role Do Moderator Prompts Play?

Moderator prompts help facilitate discussion when participants provide brief answers, disagree, introduce an unexpected topic, or remain silent.

For example, after one participant describes a concern about AI-generated assessment materials, the moderator might ask: "Have others encountered something similar or had a different experience?"

This invites additional perspectives without requiring participants to agree.

Other prompts may encourage clarification, return the discussion to the research topic, or provide opportunities for quieter participants to contribute.

AI can help develop focus group moderator prompts, but their use requires attention to group dynamics and participant comfort.

A moderator who asks every prepared prompt may disrupt the interaction rather than support it.

How Can Group Dynamics Affect the Discussion?

Focus group data are influenced not only by the questions but also by how participants interact.

Some participants may speak more frequently, while others may hesitate to disagree. Differences in professional seniority, institutional authority, or familiarity with the topic may affect who contributes and what is discussed.

For example, instructors may be reluctant to criticize institutional AI policies when senior administrators are present in the same group.

AI can suggest strategies for encouraging participation, but it cannot determine how a particular group will behave.

Researchers should consider group composition during recruitment and plan moderation procedures that allow different perspectives to be expressed.

Morgan's (1997) methodological work on focus groups emphasizes the importance of group interaction and research design decisions in shaping the data obtained.

Should Participants With Different Roles Be Placed in the Same Group?

There is no universal requirement that focus group participants have identical characteristics or professional positions.

Mixed groups may provide opportunities to explore contrasting perspectives and interactions across roles. However, differences in authority or social position may inhibit open discussion.

For example, a mixed group of instructors and administrators might reveal how institutional expectations are interpreted differently. Yet the presence of supervisors could discourage some instructors from discussing sensitive experiences.

Researchers should determine whether the advantages of mixed-group interaction outweigh the potential limitations for the particular study.

Separate groups may be more appropriate when power relationships, confidentiality concerns, or differing experiences could substantially affect participation.

AI can help compare possible group compositions, but the decision requires contextual and ethical judgment.

How Many Questions Should a Focus Group Guide Contain?

A focus group guide should provide sufficient direction without overloading the discussion.

Unlike an individual interview, a focus group involves multiple participants who may respond to one another. A single question can therefore generate an extended discussion.

Asking too many main questions may limit opportunities for interaction and encourage the moderator to move prematurely between topics.

There is no universally appropriate number of questions. The guide's length should reflect the research scope, group size, anticipated duration, and complexity of the topics.

AI can propose a manageable structure, but the actual timing should be examined through pilot discussion or other appropriate preparation.

Sequence Questions Without Controlling the Conversation Too Tightly

AI can suggest a sequence that begins with accessible questions, moves toward the central research topics, and concludes with reflection.

However, participants may introduce important issues earlier than expected. The moderator should be able to explore relevant discussion without rigidly following the prepared order.

Question order can also influence which perspectives become salient. Beginning with several questions about AI-related misconduct may frame later discussion around risks even when the study intends to explore a broader range of experiences.

Researchers should consider question sequencing principles while recognizing that group discussion requires additional flexibility and attention to interaction.

What Ethical Issues Are Particular to Focus Groups?

Focus groups raise confidentiality considerations that differ from those of individual interviews.

The researcher may establish procedures for protecting recordings and transcripts, but cannot guarantee that participants will keep everything discussed confidential after leaving the group.

Participants should therefore be informed about the limits of confidentiality and the expectations for respectful discussion before participating.

Researchers should also consider whether the questions invite disclosure of identifiable colleagues, students, institutional disputes, or sensitive professional experiences.

For example, asking participants to name colleagues who misuse AI may create unnecessary confidentiality and professional risks. A more appropriate discussion might focus on general experiences or institutional practices without requesting identifiable allegations.

AI can help flag potentially sensitive questions, but it cannot replace ethical review or determine the risks associated with a particular group composition.

Can AI Predict How Participants Will Interact?

AI can simulate a hypothetical focus group conversation and generate examples of agreement, disagreement, interruptions, or dominant participants.

Such simulations may be useful for moderator preparation or testing possible responses to difficult situations.

However, they do not establish how actual participants will interact. A model-generated conversation reflects the system's simulation rather than observed group dynamics.

Researchers should not report simulated interactions as pilot focus group findings or treat AI-generated exchanges as qualitative data collected from the intended population.

Real group discussions may reveal unexpected interpretations, interpersonal dynamics, and contextual details that the simulation did not anticipate.

Why Pilot Testing a Focus Group Guide Matters

A guide may appear coherent during review but function differently when several participants respond to the same questions.

Some prompts may generate little discussion, while others may produce lengthy exchanges that exceed the planned time. Participants may misunderstand terminology, and the moderator may discover that certain questions encourage sequential individual answers rather than interaction.

Pilot testing or an appropriate practice discussion can help identify these issues and support revisions to the guide.

Researchers should also consider moderator preparation, recording arrangements, and the practical management of participation.

AI can assist with preliminary review, but respondent-based pretesting provides information about how questions function during actual interaction that simulations cannot establish.

Document AI Assistance Without Confusing It With Group Evidence

If AI materially contributes to focus group guide development, researchers should document its role according to applicable institutional and publication requirements.

For example, AI may have been used to generate candidate questions, organize discussion topics, or identify potentially leading wording.

These activities should be distinguished from the actual focus group discussions and the qualitative analysis of participant-generated data.

Researchers should not claim that AI-generated questions were validated through simulated focus groups or that the model predicted the themes participants would discuss.

The quality of the resulting study depends on the appropriateness of the research design, the conduct of the discussions, and the interpretation of the actual data.

Using AI to Create an Interaction-Oriented Focus Group Guide

The revised guide does not simply ask each instructor to provide an isolated opinion. It creates opportunities for participants to respond to one another and explore differences in how institutional expectations are experienced.

The researcher must still remain attentive to group dynamics, including whether some participants dominate the discussion or others appear reluctant to contribute.

Common Mistakes When Using AI to Create Focus Group Guides

Design the Guide for Interaction, Not Just Individual Responses

When using AI to prepare a focus group discussion guide, begin by identifying why group interaction is valuable for the research question.

Ask AI to suggest questions that invite participants to describe experiences, compare perspectives, and respond to ideas introduced by others. Then evaluate whether the questions are appropriate for the participants and research context.

Review the guide with the moderator and pilot the discussion where appropriate. During the session, remain responsive to participants rather than requiring the conversation to follow every prepared question.

AI can help prepare the discussion, but the methodological value of a focus group depends substantially on what participants contribute through their interaction.

Before Using an AI-Generated Focus Group Discussion Guide

Frequently Asked Questions About AI-Generated Focus Group Guides

AI Can Prepare the Discussion Guide, but Participants Create the Interaction

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.