The research interview looks like a conversation and is judged like an instrument. Examiners can't watch you interview — they infer your rigour from the protocol, the audit trail and the depth of your quotes. Depth doesn't come from charisma; it comes from preparation, disciplined listening and good probes.
Choose your structure deliberately
- Structured — fixed questions, fixed order; essentially an oral survey. Use only when comparability across many interviewers matters.
- Semi-structured — a guide of open questions with freedom to probe and reorder. The doctoral default, and what the rest of this guide assumes.
- Unstructured / narrative — a grand-tour opening and follow-the-participant logic; powerful for phenomenology and life-history work, demanding for novices.
Build the interview guide from your research questions
- 1Map each research question to 2–4 interview topics — this matrix goes in your appendix and answers the viva question 'how does your instrument cover your questions?'
- 2Draft open, experience-near questions: 'Walk me through the last time…' beats 'Do you think that…?' every time.
- 3Sequence from easy and concrete to sensitive and abstract; never open with the hardest question.
- 4Prepare planned probes under each question: 'Can you give me an example?' · 'What happened next?' · 'How did that feel?'
- 5Pilot with 1–2 participants and revise — pilots routinely kill your favourite question and reveal the one you forgot. Start from our interview guide template.
In the room: the craft
- Consent first, recorder second — informed consent (use a proper consent form) covering recording, storage and withdrawal, before anything is taped.
- Tolerate silence. Count three seconds before rescuing a pause — the best material lives on the far side of silence.
- Probe, don't lead. 'Tell me more about that' is a probe; 'So you found it frustrating?' plants your word in their mouth.
- Track, don't interrupt. Note follow-ups on your guide and return to them at a natural break.
- Close deliberately: 'Is there anything important I haven't asked about?' — the question that yields findings you didn't know to seek.
How many interviews? Saturation, honestly used
Most doctoral studies land at 12–25 semi-structured interviews, but the defensible answer is saturation: the point where new interviews stop yielding new codes or themes. Document it — a simple table of 'new codes per interview' trending to zero is far stronger viva evidence than citing a rule of thumb.
After the interview
Write a memo within 24 hours (context, surprises, emerging hunches), transcribe verbatim, and anonymise with a participant-ID key stored separately. Then the analysis begins — usually thematic analysis, with NVivo earning its keep above roughly ten transcripts.
Video interviews are now standard doctoral practice — just address them in your method: platform, recording consent, handling of connection failures, and any rapport limitations. The craft above transfers intact.
Our qualitative research mentoring covers protocol design, pilot review and mock interviews — the highest-yield practice hours a qualitative scholar can buy.
Frequently asked
How long should a research interview be?+
45–60 minutes for most semi-structured doctoral interviews. Under 30 minutes rarely reaches depth; beyond 90, quality drops for both parties. Tell participants the expected length in the consent process and respect it.
Should I transcribe interviews myself?+
At least some — transcription is a first pass of analysis, and immersion is where early themes surface. AI transcription tools can produce the draft, but you must correct them against the audio, and their use on consented data should be cleared with your ethics process.
How do I interview in Hindi or a regional language?+
Interview in the language the participant thinks in — depth beats convenience. Transcribe in the original language, analyse as close to the original as possible, and translate only the quotes you present, noting the translation procedure in your method section.
phdguide's mentors are senior academics, former supervisors, statisticians and publication specialists with 25+ years' average experience guiding MBA, MPhil and PhD scholars from topic to viva.
Ethical, compliant guidance: We provide academic support, mentoring, analysis, editing and structuring — not authorship. Your work stays compliant with university policies.