FGDs done badly produce groupthink transcripts. FGDs done well surface the "why" behind the numbers. This pack covers recruitment, facilitation, consent, and rapid thematic analysis -- so your FGDs produce insights, not noise.
4 modules~100 minInteractive -- India-context
Your progress
0% complete
Your Capstone
Complete FGD Facilitator's Guide
Recruitment criteria, facilitator guide with questions and probes, consent script, logistics checklist, and rapid analysis template -- ready for your next field study.
Module 1 -- ~25 min
Recruitment and group composition
The quality of an FGD is determined before it begins -- by who is in the room. Bad recruitment produces bad data. The two cardinal sins: (1) letting a programme coordinator hand-pick "good" participants, and (2) mixing people with power differentials in the same group.
Homogeneous on power, heterogeneous on experience. Same gender, similar age range, similar social position. Different experiences with the programme.
No authority figures. If the sarpanch is in the room, nobody else speaks freely. If the teacher is present, students will not critique the programme.
Separate groups for separate perspectives. Men and women separately. Programme staff and beneficiaries separately. Different caste groups if caste dynamics are relevant.
India-specific recruitment challenges
In Indian villages, the default "community meeting" recruits the already-powerful: men, upper-caste, landowners, those close to the sarpanch. Active recruitment of marginalised voices requires: (1) visiting homes, not calling meetings; (2) scheduling around women's work patterns (not morning, not cooking time); (3) providing childcare or allowing children; (4) choosing a venue where Dalit and Adivasi participants feel comfortable -- often not the panchayat bhavan.
Your Recruitment Plan
How many groups? Who is in each? What are you keeping homogeneous?
How will you actually find and invite participants? Avoid programme staff hand-picking.
Saved
Self-check
You are running an FGD about a girls' education programme. The school principal offers to sit in "to help with translation." Should you accept?
Yes -- the translation help is valuable
No -- the principal is an authority figure; students will self-censor. Hire a separate local-language interpreter.
Yes, but ask the principal not to speak
Depends on the principal's personality
Correct. Authority dynamics override good intentions. Even a silent principal changes what students say. Budget for an independent interpreter who is not affiliated with the school or programme.
Module 2 -- ~25 min
Writing the facilitator guide
A facilitator guide is not a list of questions. It is a structured conversation plan with opening rituals, warm-up questions, core exploration questions, probes, transition phrases, and closing. The guide ensures consistency across multiple FGDs while leaving room for emergent themes.
Guide structure (60-90 minute FGD)
Opening (5 min) -- introductions, ground rules ("everyone's view matters," "no right answers," "what is said here stays here"), consent reminder
Warm-up (5 min) -- easy, non-threatening question. "Tell us your name and one thing you enjoy about living in this village."
Introductory question (5 min) -- broad, open. "When you hear '[programme name],' what comes to mind?"
Core questions (30-45 min) -- 4-6 key questions with probes. These are the questions that answer your research question.
Summary and verification (5 min) -- "I heard three main themes: X, Y, Z. Did I get that right? Is there anything we missed?"
Closing (5 min) -- "Is there anything else you want to share?" Thank participants. Provide refreshments.
Worked example -- Core question with probes
Core Q3: "Some women continued with the programme and some stopped. From your experience, what made the difference?"
Probe a: "Can you give a specific example of something that made it hard to continue?"
Probe b: "Was there anything about the timing or location that affected attendance?"
Probe c: "How did your family respond to your participation?"
Your Facilitator Guide -- Core Questions
Write each question + 2-3 follow-up probes
Saved
Self-check
Your facilitator guide has 12 core questions for a 60-minute FGD. What is likely to happen?
All questions will be covered with depth
You will rush through questions, lose depth, and exhaust participants -- cut to 4-6 core questions with probes
The facilitator will use judgement to select the best ones
Extend the FGD to 2 hours
Correct. With opening, warm-up, and closing, you have ~40 min for core questions. 12 questions = 3.3 min each = survey-style rapid answers, not discussion. Cut to 4-6 questions with probes. Depth beats breadth in qualitative research.
Module 3 -- ~25 min
Recording, consent, and logistics
Logistics determine whether good FGD design translates into usable data. Three things go wrong most often: consent is done perfunctorily, the recording fails, and the venue is wrong.
Consent protocol
Read the consent script in local language -- slowly, clearly
Explain: purpose, who will hear the recording, how data will be used, that they can leave any time
For non-literate participants: verbal consent with witness, recorded on audio
Separate consent for audio recording -- some may consent to participate but not to recording
For children/adolescents: parental consent + child assent. Both required.
Recording and note-taking
Two recording devices -- one phone + one dedicated recorder. Batteries full. Test before starting. One will fail; two rarely do.
Note-taker -- a second person captures who speaks, body language, group dynamics, seating arrangement. The transcript alone misses half the data.
Participant IDs -- assign P1-P10, noted on the seating chart. Never use names in transcripts.
Venue matters more than you think
In rural India, venue signals power. The panchayat office favours dominant-caste speakers. The programme office implies the "right" answers are pro-programme. The best FGD venues are neutral community spaces (school verandah after hours, temple/mosque courtyard if appropriate, an SHG member's home for small groups). Check: Is it accessible? Private enough? Not too noisy? Has shade?
Your Logistics Checklist
Written or verbal? Who witnesses? How is recording consent handled?
Facilitator, note-taker, travel, refreshments, venue, participant transport
Saved
Self-check
Your audio recorder fails 20 minutes into the FGD. You have detailed notes from the note-taker. Can you use this data?
No -- without audio, the data is invalid
Yes, partially -- detailed notes are usable for thematic analysis, though you lose verbatim quotes and should document the limitation
Yes, fully -- notes are just as good as audio
Restart the FGD from the beginning
Correct. Detailed notes are usable but weaker than transcripts -- you lose exact wording and may miss quieter voices. Use the data with a limitation statement. This is why you should always have two recording devices.
Module 4 -- ~25 min
Rapid thematic analysis
You have 4-6 FGD transcripts. Now extract the findings. "Rapid thematic analysis" is a structured approach that gives you defensible themes without requiring NVivo or 3 months of coding. It works well for evaluation-scale qualitative data.
The 5-step rapid process
Debrief immediately -- within 2 hours of each FGD, facilitator + note-taker discuss: what were the main themes? What surprised us? What contradicted other FGDs?
Read all transcripts once without coding. Just absorb. Note initial impressions.
Create a coding framework from your research questions + emergent themes. 8-15 codes is typical.
Code using a matrix: rows = codes, columns = FGD groups. For each code, extract key quotes and note agreement/disagreement across groups.
Write theme statements -- each theme is a finding expressed as a sentence: "Women who dropped out consistently cited husband's opposition, not lack of interest."
Your Analysis Plan
From your research questions + anticipated themes
Which groups will you compare? What differences do you expect?
Saved
Self-check
Two FGDs strongly agree that the programme helped. One FGD strongly disagrees. What should you report?
Report the majority view -- 2 out of 3 is a finding
Report all three and analyse why the third group differs -- the disagreement is often the most important finding
Run a fourth FGD to break the tie
Average out the responses
Correct. Qualitative research does not work by majority vote. The dissenting group may be the dropout group, or a different caste, or from a different geography. The disagreement often points to the most actionable finding: "The programme works for X but not for Y because of Z."
Capstone
Your FGD Facilitator's Guide
Click Build my brief to compile your complete facilitator's guide.