How to Conduct Research Interviews

A research interview looks like a conversation, but every part of it — from the first question to the way you end it — is a methodological decision you will need to justify.

Interviews are one of the most common methods in student qualitative research, and one of the most underestimated. It is easy to assume that talking to people is straightforward. The difficulty becomes clear when transcripts come back full of short answers, leading questions and tangents that do not address the research question. Good interview data is designed long before the first conversation.

Choose the right interview format

Interviews range from highly structured, where every participant gets the same questions in the same order, to unstructured conversations around a theme. Most student projects use semi-structured interviews: a guide of core questions with freedom to follow up on what participants raise. This balances comparability across interviews with the depth that makes qualitative data valuable.

State and justify your choice in the methodology. The format should follow from your research question and analytic approach.

Write an interview guide from your research question

Begin with your research question and break it into three to five topic areas. Under each, write one or two main questions and a few possible prompts. A typical guide for a forty-five-minute interview has eight to twelve main questions — more than that and you will rush. Order them from easier, descriptive questions to more reflective or sensitive ones.

Ask open, neutral questions

Question wording shapes answers. Closed questions produce yes or no. Leading questions signal the answer you hope for. Double-barrelled questions ask two things at once and get half an answer.

Questions beginning “how,” “what” or “tell me about” usually work best. “Why” questions can feel confrontational, so soften them where possible: “What led you to that decision?”

Pilot the guide

Run one or two practice interviews with someone similar to your participants. Piloting reveals questions that confuse people, sections that run long and topics you forgot. It also lets you practise with your recording equipment. Revise the guide afterwards and mention the pilot in your methods chapter — examiners see it as evidence of care.

Get ethics right before recruiting

Interview research almost always needs ethical approval. You will need a participant information sheet explaining the purpose, what taking part involves, how data will be stored and the right to withdraw, and a consent form. Think carefully about anonymity: in small or specialist populations, quotes can identify people even without names. Plan how you will handle distress if your topic is sensitive.

Recruit purposefully

Qualitative samples are chosen for the information they can provide, not for statistical representativeness. Define your inclusion criteria clearly and recruit people who can speak to your question. Common approaches include purposive sampling, snowball sampling through existing participants, and gatekeepers in relevant organisations. Many student projects use eight to fifteen interviews, but the right number depends on the depth of each and on whether new interviews are still adding insight.

Run the interview well

Start with a brief explanation, confirm consent and check the recording is working. Begin with an easy question to build rapport. During the interview, listen more than you talk, allow silences — participants often fill them with their most useful reflections — and follow up on vague answers with prompts such as “could you give me an example?”

Handle difficult moments calmly

Not every interview goes smoothly. Some participants give very short answers; prompts such as “what happened next?” or “how did that feel at the time?” usually help, as does reassuring them there are no right answers. Others talk at length on tangents; gently steer back with “that’s really helpful — could we come back to…”. If a participant becomes upset, pause, ask whether they want to continue, and respect their decision. Never push for information someone seems reluctant to share.

Online versus in-person interviews

Video calls widen your reach, lower costs and suit busy participants, but they bring connection problems, distractions and sometimes less rapport. Telephone interviews remove visual cues entirely but can make some participants more comfortable discussing sensitive topics. Whatever the format, test the technology beforehand and have a backup recording method.

Transcribe and manage the data

Transcribe as soon as possible after each interview, while the conversation is fresh. Automated transcription saves time but always needs checking against the recording. Anonymise as you go, replacing names and identifying details with codes or pseudonyms, and store files securely as promised in your consent forms. Writing a short reflective memo after each interview helps your later analysis.

Prepare for analysis while you collect

Analysis should not wait until every interview is finished. Reading early transcripts and noting emerging ideas lets you refine prompts for later interviews and recognise when new conversations are adding little that is new. Decide early which analytic approach you will use — thematic analysis, interpretative phenomenological analysis or framework analysis, for instance — because each has implications for how you transcribe, how many interviews you need and how you report findings.

Reflect on your own position

As the interviewer, you influence what participants say — through your questions, your reactions and who you are. Qualitative research does not ask you to eliminate this, which is impossible, but to acknowledge it. A short reflexivity section in your methodology, describing your relationship to the topic and participants and how you managed it, strengthens the credibility of your findings.