Use when designing and facilitating focus groups for qualitative data collection.
Focus groups generate data through interaction: participants respond to prompts, build on each other’s ideas, and sometimes reveal normative assumptions, vocabularies, and disagreements that individual interviews suppress or never surface.
Use focus groups when group sense-making, consensus/dissensus, or cultural scripts are central—e.g., testing how a community talks about a stigma-laden topic, exploring how teams coordinate, or mapping shared meanings of a policy. Use individual interviews when sensitive disclosures, power asymmetries within a peer group, or detailed personal narratives are primary. Avoid focus groups when participants may fear retaliation from others in the room.
Aim for homogeneity on the dimension that drives comfort (e.g., job role, experience level) and careful heterogeneity only when cross-perspective friction is analytically desired. Typical size is 4–8 participants; smaller groups suit complex tasks or vulnerable topics. Over-recruit slightly to handle no-shows.
Structure sessions with: (1) welcome, consent, ground rules (respect, confidentiality limits, one speaker at a time); (2) warm-up; (3) 3–5 focused question blocks with planned follow-ups; (4) closing reflection. Questions should invite concrete examples (“Tell us about a time…”) before abstract generalizations (“What do people think about…?”).
Track who speaks, who is silent, who aligns, who challenges. Silence can mean agreement, fear, or disengagement—follow up ethically. Manage dominant speakers kindly (“Let’s hear from someone who hasn’t spoken yet”). Note laughter, interruptions, and collective corrections; these are analytic material about norms, not mere noise.
Transcribe by speaker turn with minimal identifiers in transcripts used for analysis. Code both content (themes) and interaction (alignment episodes, storytelling chains). Compare focus group findings with interviews or observations when possible to avoid over-weighting performative public accounts.
Use multiple microphones when affordable; backup recording when ethical and practical. Obtain explicit consent for recording and clarify that confidentiality cannot be guaranteed among participants—set ground rules accordingly.
There is no magic N. Plan iterative rounds until you observe pattern repetition and diminishing novelty for your analytic goals (analogous to saturation thinking). Reporting should justify group count by data quality and stability of themes, not by a fixed rule.
Treat each exchange as comparable incidents. The moderator should avoid imposing preconceived categories; probe emergent distinctions participants introduce. Memo immediately after each session on hypotheses about categories, conditions, and consequences suggested by interaction patterns.
Do not coerce disclosure; debrief emotionally heavy sessions; offer referrals when appropriate. For industry or workplace groups, clarify whether managers can attend and how that affects candor.
Opening script essentials: purpose, voluntary participation, confidentiality limits among peers, recording consent, respectful turn-taking.
Moderation moves: “What do others think?” after a strong opinion; invite concrete stories before general norms; manage dominators without shaming.
Analysis reminder: code both what was said and how the group negotiated agreement, humor, silence, or conflict.