Blog

· By Thesisroom

Semi-Structured Interview Guide: How to Build One, With a Template

What a semi-structured interview is, how to turn research questions into main questions and probes, how to run and record each interview, and what to report in your methods.

This guide has a printable version: open the printable semi-structured interview guide template

Two people sit across a white table holding gray mugs during a quiet one-to-one conversation
Photo by Priscilla Du Preez 🇨🇦 on Unsplash

A semi-structured interview is a qualitative interview “organized around a set of predetermined open-ended questions, with other questions emerging from the dialogue” between interviewer and participant (DeJonckheere and Vaughn, 2019, quoting DiCicco-Bloom and Crabtree). A semi-structured interview guide is the written list of those planned questions, with the follow-up questions and probes for each one, that you take into every interview.

Below: how the format differs from structured and unstructured interviews, eight steps to build the guide, how to run and record each interview, and what to report in your methodology chapter. There is an invented worked example and a template you can print.

What is a semi-structured interview, compared with structured and unstructured ones?

The difference is how much you fix in advance. Harvard Library’s qualitative research guide, quoting SAGE encyclopedia entries, places the semi-structured interview between the other two (Harvard Library).

In a semi-structured interview, the researcher “has more control over the topics of the interview than in unstructured interviews”. Unlike structured interviews or questionnaires that use closed questions, “there is no fixed range of responses to each question”. Unstructured interviews ask “relatively open-ended questions” to learn how participants see the topic.

StructuredSemi-structuredUnstructured
QuestionsClosed, like a questionnairePredetermined but open-endedRelatively open-ended
AnswersA fixed range of responsesNo fixed range of responsesOpen-ended

DeJonckheere and Vaughn call the semi-structured interview the most common type of interview in qualitative research and in healthcare. They note that the interviewer typically changes the sequence and wording of questions “to best fit the interviewee and interview context”. That flexibility is the point of the format, and it is why the guide matters so much.

When should a thesis use semi-structured interviews?

Use them when your research question asks how or why people experience, decide or understand something. DeJonckheere and Vaughn say the method works when the researcher wants:

  • Open-ended data: qualitative answers in the participant’s own words.
  • Thoughts, feelings and beliefs: what participants think and feel about a particular topic.
  • Depth on sensitive issues: personal and sometimes sensitive issues, explored in depth.

They also list other uses: collecting new exploratory data, triangulating other data sources, and checking findings with participants. In a mixed methods thesis, interviews can explore new concepts in a qualitative phase or explain results from a quantitative one.

If your question needs every participant to choose from the same answer options so you can count them, a structured questionnaire fits better. That follows from the definitions above: only the structured format has a fixed range of responses.

How do you build a semi-structured interview guide, step by step?

Work down from your research questions to individual questions, then test the guide before real data collection. Two published frameworks set out the order.

Kallio and colleagues (2016) ran a systematic methodological review and propose five phases. They are identifying the prerequisites for using semi-structured interviews, retrieving and using previous knowledge, formulating the preliminary guide, pilot testing it, and presenting the complete guide. Castillo-Montoya (2016) proposes four phases: aligning interview questions with research questions, constructing an inquiry-based conversation, receiving feedback on the protocol, and piloting it.

The eight steps below combine both frameworks with the practical detail in DeJonckheere and Vaughn.

  1. Confirm that interviews fit your research question. Kallio’s first phase is identifying the prerequisites for using semi-structured interviews. DeJonckheere and Vaughn note that qualitative research questions typically start with “What”, “How” or “Why” and explore one concept from participants’ perspectives. If your research questions and objectives do not yet match, line up your objectives and research questions first.
  2. List what you already know. Kallio’s second phase is retrieving and using previous knowledge. Pull the topic areas out of your literature review and framework, and write each one as a short topic, not yet a question.
  3. Map every research question to two or three main questions. Castillo-Montoya’s first phase is alignment. Draw a grid with research questions across the top and interview questions down the side. A question that serves no research question gets cut, and a research question with no interview question is a gap.
  4. Write one opening question and five to 10 core questions. DeJonckheere and Vaughn describe a “grand tour” opening question: a general question about something the participant knows well, to get them talking. Their core questions number five to 10, relate directly to what you need to know, and are typically asked of all participants.
  5. Check every question’s wording. DeJonckheere and Vaughn say all questions should be open ended, neutral and clear, avoid leading language, use familiar words and avoid jargon. Read each question aloud and ask whether a participant could answer it with “yes” or “no”.
  6. Add planned follow-up questions and probes. Planned follow-ups ask for detail on parts of a core question. Unplanned ones arise from what the participant says. DeJonckheere and Vaughn call follow-up questions “equally important” to the core questions.
  7. Order the questions from easy to hard. Most interviews, they write, start with an easy, context-setting question before moving to more difficult or in-depth ones. Put the most sensitive question after rapport is built, and end with an open closing question.
  8. Get feedback, pilot the guide, and keep revising. Castillo-Montoya’s third and fourth phases are feedback and piloting. DeJonckheere and Vaughn say a pilot helps you pace the interview, keep a conversational tone and adjust questions. Because collection and analysis run together, weak questions can be replaced and new probes added later, so date each version of the guide.

Which probes can you plan in advance?

Probes are short moves that get a participant to say more. Five of the six technique names and their descriptions follow DeJonckheere and Vaughn’s table of probing techniques, and the sixth is renamed. The examples are ours, except “Yes... okay...” and “Tell me how you came to that decision”, which are theirs.

TechniqueWhat you doExample
Wait timeStay silent after a question until the participant speaks(Ask, then wait)
EchoRepeat or summarize their words“So the ward went quiet after two.”
Verbal agreementShort affirming words“Yes... okay...”
ExpansionAsk for more detail or an example“Can you give me an example of that?”
ExplanationAsk what a word or phrase means to them“What do you mean when you say you ‘held off’?”
ReasoningAsk how they reached a decision (the source calls this “leading”)“Tell me how you came to that decision.”

The reasoning probe is not a leading question in the sense of steering an answer. It asks the participant to explain a decision they have already described.

What does a finished guide look like? A worked example

This example is invented for illustration. A Master’s candidate in nursing asks: “How do newly qualified nurses on night shifts decide whether to escalate a concern about a patient?” The plan is to interview twelve nurses from two hospitals.

The literature review gave four topic areas: confidence, ward hierarchy, past experience and support from senior staff. Each core question below serves at least one of them.

  1. Opening (grand tour): “Can you walk me through a typical night shift on your ward?”
  2. Core question 1: “Tell me about a recent night when you were worried about a patient.” Probes: “What did you notice first?” and “What did you do next?”
  3. Core question 2: “How did you decide whether to call someone?” Probes: “Who did you call, and why that person?” and “Tell me how you came to that decision.”
  4. Core question 3: “What makes it easier or harder to escalate a concern at night?” Probe: “Can you give me an example of that?”
  5. Core question 4: “Tell me about a time you decided not to escalate.” Probe: “Looking back, how do you see that decision now?”
  6. Core question 5: “What would you tell a nurse starting nights next month?”
  7. Closing: “Is there anything about escalating at night that we have not covered?”

Core question 4 is the most sensitive, so it comes after three questions that build rapport. The first draft of the guide had three weak questions. Here is how each was fixed.

First draftProblemRevised
“Don’t you think the hierarchy stops nurses from calling doctors?”Leading, and answerable with yes or no“What, if anything, makes it hard to call a doctor at night?”
“Do you use the early warning score?”Closed“How does the early warning score fit into your decision?”
“How would you rate your escalation self-efficacy?”Jargon“How confident do you feel when you pick up the phone, and what affects that?”

How do you run and record a semi-structured interview?

Settle consent and logistics before the interview, listen more than you talk during it, and write a memo straight after. The steps below follow DeJonckheere and Vaughn unless another source is named.

  1. Tell participants what to expect. They should know the length, the purpose, why they were selected and who will be there. Tell them they can refuse to answer questions or withdraw at any time, including during the interview.
  2. Ask about recording. DeJonckheere and Vaughn recommend audio recording so you can concentrate and build rapport instead of taking extensive notes. Participants should know it is used for data collection and that they can refuse to be recorded.
  3. Pick and test the place. Choose somewhere quiet and private where you can get a clear recording, and test your equipment there. If you are the sole interviewer with no note-taker, they recommend two recording devices in case one fails.
  4. Open the interview properly. Briefly explain the study, the consent procedure, why you chose this person and the format of the interview. Check the recorder is running.
  5. Listen actively. Keep a warm, conversational tone, not a rote question-and-answer pattern. Silence, nods and short utterances encourage people to say more, and brief notes help you track which questions are covered.
  6. Write a memo after each interview. A memo records what you are learning from the data. DeJonckheere and Vaughn add that memos also improve later interviews.
  7. Transcribe and organize. They note that most researchers will want verbatim transcripts so they can use participants’ exact words. The UK Data Service’s deposit guidance also asks for consistent formatting with header metadata in each file, described transcription conventions where applicable, and identifiers removed or replaced where necessary.

The same UK Data Service page asks for a spreadsheet summarizing each interview, with its ID, context, topics covered and file names. That log helps even if you never deposit your data. When you quote participants in the thesis, the rules are different from citing a source: see how to cite an interview in a thesis.

Analysis starts while you are still interviewing. DeJonckheere and Vaughn say the right approach depends on your research question and qualitative design, so name it in your methods, not here.

How do you describe semi-structured interviews in the methodology chapter?

Report why you chose the format, how you built and tested the guide, and how you ran, recorded and transcribed the interviews. Kallio and colleagues suggest researchers consider their five phases and “justify the decisions made during it”.

For the justification itself, our guide to writing a methodology chapter shows how to name the choice, the nearest alternative, the reason and the consequence. For the interview procedure, cover these points:

  • Why this format: why semi-structured interviews and not structured ones, unstructured ones or focus groups, tied to your research question.
  • The guide: how you developed it, how it maps to your research questions, how many core questions it has, and which appendix holds it.
  • Feedback and piloting: who reviewed or tested the guide, and what changed as a result.
  • Changes during collection: any questions added, dropped or reworded, and when.
  • Participants and setting: how participants were chosen (DeJonckheere and Vaughn describe purposeful sampling) and where the interviews took place.
  • Conduct and recording: who interviewed, in person or online, how consent was taken, how the interviews were recorded and whether you took field notes.
  • Transcription: verbatim or not, the conventions you used and how you anonymized names.

If your field uses reporting checklists, look at COREQ (Tong, Sainsbury and Craig, 2007). It is a 32-item checklist for interviews and focus groups, grouped into three domains: research team and reflexivity, study design, and data analysis and reporting.

Here is an invented example of the procedure paragraph for the nursing study above:

Data were collected through semi-structured interviews with twelve newly qualified nurses. The guide (Appendix C) had one opening question and five core questions with planned probes, developed from the literature review and mapped to both research questions. It was reviewed by the supervisory team and piloted with one nurse outside the sample, after which one leading question was reworded.

Interviews took place in a private room at each hospital, were audio recorded with consent and were transcribed verbatim, with names replaced by participant codes. Brief field notes were taken during each interview and a reflective memo was written afterward.

Write a proposal version of this in the future tense, as the guide to research methodology for a research proposal explains. Once the chapter is drafted, Thesisroom’s Chapter Audit leaves margin comments on it, the way a careful supervisor marks a chapter.

If your examiners press on sample size, trust or bias, defending qualitative research in a viva has wording for those questions.

What mistakes weaken a semi-structured interview study?

The main ones are thin probing, poor listening, a weak guide, insensitive questions and underestimating the work. DeJonckheere and Vaughn list these as common problems, and each has a fix you can plan for.

  • Not probing: answers stay thin. Write planned probes under every core question and practice the wait-time probe.
  • Not listening actively: you miss what to follow up. Take short notes on phrases to come back to.
  • A weak guide: closed or leading questions produce short answers. Run every question through step 5 above.
  • Insensitive questions: participants close down. Place sensitive questions late and word them neutrally.
  • Underestimating the work: recruiting, interviewing, transcribing and analyzing all take resources. Plan time for each in your thesis timeline.

Printable semi-structured interview guide template

Fill in one copy per interview and keep it with the recording and your memo.

FieldYour entry
Study title
Research question(s)
Participant code
Date
Place or platform
Interviewer
Guide version
Consent and recording confirmed
SectionMain questionPlanned probesResearch question it servesCovered
Opening
Core 1
Core 2
Core 3
Core 4
Core 5
Core 6
Closing

Before the interview:

  • Participant knows the purpose, length and why they were chosen
  • Participant knows they can skip questions or withdraw at any time
  • Consent to record is confirmed
  • Recorder tested at the location, with a backup device
  • Quiet, private room booked

After the interview:

  • Recording saved under the participant code
  • Memo written on content and process
  • Phrases worth probing in later interviews noted
  • Any change to the guide dated and logged
  • Interview added to the interview log

Frequently asked questions

How many questions should a semi-structured interview guide have?DeJonckheere and Vaughn describe five to 10 core questions that relate directly to what you need to know, plus an opening question and planned follow-ups.
Do I have to ask every question in the same order?No. In a semi-structured interview the interviewer typically changes the sequence and wording to fit each participant and context. Core questions are typically asked of all participants, so make sure each one is covered by the end.
Can I change my interview guide after I start interviewing?Yes. DeJonckheere and Vaughn note that collection and analysis happen together, so weak questions can be replaced and probes added. Date each version, report the changes in your methodology chapter and check whether your ethics approval covers them.
Is a semi-structured interview qualitative or quantitative?It is a qualitative data collection method with predetermined but open-ended questions and no fixed range of answers. In a mixed methods study it can form the qualitative phase, exploring new concepts or explaining quantitative results.
How many participants do I need for semi-structured interviews?It depends on your purpose and question. DeJonckheere and Vaughn say a family medicine project can be meaningful with as few as 8 to 12 participants, and cite information power as one way to decide. Rules and expectations differ between universities, so agree a rationale with your supervisor.
Should I record semi-structured interviews?Audio recording is recommended so you can concentrate on the participant instead of taking extensive notes. Tell participants the recording is for data collection and that they can refuse it. If you interview alone with no note-taker, use two recording devices.
Does the interview guide go in the thesis appendix?Presenting the complete guide is the last of Kallio and colleagues’ five phases, and an appendix lets examiners see your exact questions. Check your university’s rules on what appendices may contain.

Sources

How this guide was made: Thesisroom drafted this guide with the help of AI writing tools. Every fact, figure and source in this post was checked against its original source before publishing.