The most common mistake in qualitative fieldwork is treating an interview guide as a script and reading it out in order. That produces short answers to your questions rather than an account of the participant's experience — a questionnaire administered slowly.
Enter your topic and this tool builds a funnel-ordered guide: rapport-building openers, core questions tied to your objectives, a bank of probes to use on any answer, and closing questions. The technique notes matter as much as the questions.
Opening questions are broad and easy to answer: role, background, how the topic arises in the participant's life. Their purpose is rapport, not data. Someone who has spent two minutes talking comfortably will answer the difficult question at minute twenty far more fully than someone hit with it immediately.
Core questions narrow toward your objectives. Closing questions widen again — "what would you change?", "what have I not asked about?" — because the last question routinely surfaces the thing you did not know to ask. That final question has rescued more qualitative projects than any other single technique.
"Could you say more about that?", "what did that look like in practice?", "can you give me an example?" — these turn a one-sentence answer into the concrete, specific material that makes a findings chapter readable. Keep them written down, because they are easy to forget under the pressure of a live interview.
Silence is the most underused probe. Waiting three or four seconds after an answer feels uncomfortable and very often produces the most valuable sentence in the transcript. Resist filling the pause with your next question.
"Don't you think the system is too slow?" gets your opinion confirmed. "How would you describe the speed of the system?" gets theirs. Leading questions are the most damaging error in qualitative interviewing because the data looks fine — it simply reflects you rather than the participant.
Ask one thing at a time; a double question gets one answer and you will not know which part it addressed. Keep questions open, since anything answerable with yes or no wastes a turn. And avoid your own theoretical vocabulary — ask about what people do and experience, not whether they "experience institutional barriers".
A structured interview asks fixed questions in fixed order, which makes responses comparable but forecloses anything unanticipated. An unstructured interview follows the participant almost entirely, which gives depth but makes comparison across participants hard.
Semi-structured sits between: a guide of topics you intend to cover, asked in whatever order the conversation takes, with freedom to follow what emerges. It is the usual choice for student qualitative work because it balances comparability against discovery. Name your choice in your methodology and explain why it fits your question.
Ask permission to record and wait for an explicit yes. Explain how the recording will be stored and when it will be destroyed. Keep interviews to 45–60 minutes; attention and data quality both fall away after that, for you as much as the participant.
Transcribe within a day or two while you still remember tone and context, and write a short memo immediately after each interview noting what struck you. Those memos become the beginning of your thematic analysis, and they are impossible to reconstruct weeks later.
Eight to fifteen is typical for a student project, with the aim of variety rather than volume. The usual stopping rule is saturation — when new interviews stop producing new themes.
Forty-five to sixty minutes. Beyond that, both attention and data quality decline.
Send the topics if it helps participants feel prepared, but not the full guide — rehearsed answers lose the spontaneity that makes qualitative data valuable.
A question opens a new topic; a probe deepens the answer just given. Most of your best material comes from probes.
Yes — informed consent, usually written, plus explicit permission to record. Both normally require ethical approval before you start.