Semi Structured Interviews
Semi-structured interviews produce some of the richest data in qualitative research. Unlike structured interviews that follow a fixed script, semi-structured interviews can go off-script, allowing more room to probe, clarify, and follow the conversation where it leads. While this makes the format flexible, it also means it takes more organization to analyze the data.
What is a semi-structured interview?
A semi-structured interview uses a set of open-ended questions paired with follow-up probes that let you explore responses in more depth.
Some questions are planned in advance. Others emerge from the conversation itself. This balance helps you stay focused on your research questions while leaving room for unexpected insights participants bring up.
This interview approach works especially well when you're exploring participants' experiences, beliefs, and perspectives on a specific topic. Once you start collecting data, Delve qualitative coding software makes it easier to organize raw transcripts into insights without losing track of what you’re finding.
Key characteristics
Semi-structured interviews share a few defining characteristics:
They generate open-ended, qualitative data
Interviews follow a guide, not a script
You can probe, clarify, and follow promising threads
Conversations remain focused without feeling rigid
Participant experiences and interpretations take center stage
This structure gives you consistency across interviews without flattening individual responses. You can compare what different participants say about the same topic while still capturing how each person uniquely experiences it. It’s a lot of data to wrangle but it's rich
As you code ten or fifteen transcripts, keeping all of that organized while you move between coding and building a narrative gets messy. Tools like Delve keeps that process from becoming its own problem.
When to use semi-structured interviews
Semi-structured interviews work well when you need depth over breadth. Researchers use them when understanding how or why something happens matters more than measuring how often it does.
They're a natural fit for thematic analysis, where you build meaning across multiple conversations and look for patterns in how people experience or interpret something. They're also used in grounded theory research, where the interview data itself shapes the theory you develop. For studies focused on individual lived experience, interpretative phenomenological analysis pairs well with semi-structured interviewing because the flexible format gives participants room to describe things in their own terms.
If your research question asks "what do people think or experience?" rather than "how many?" you're probably in semi-structured interview territory.
Semi-structured vs unstructured interviews
What is the difference between a semi-structured and unstructured interview? It mostly comes down to intentional structure.
Semi-structured interviews start with predetermined questions that anchor the conversation to your research topic. You get rich, detailed data, and can add probing questions as the conversation develops.
Unstructured interviews rely almost entirely on spontaneity. While this can surface unexpected ideas, it also makes it easier to drift away from your research focus and harder to compare responses across participants later during analysis.
For most qualitative studies, like those we covered earlier, semi-structured interviews strike that practical balance between structure and flexibility.
How to prepare and conduct semi-structured interviews
Think of this as a loose sequence rather than a rigid checklist.
Start with your research objective. Be clear about what you want to learn and why interviews are the right method for your study.
Design open-ended questions. Keep them simple, neutral, and flexible. Avoid leading language, especially on sensitive topics where participants might feel pressure to answer a certain way.
Identify participants. Depending on your study, you might interview a subset of a larger group or everyone in a small population. Your sampling approach depends on your methodology.
Plan the logistics. Decide how and where interviews will happen. Obtain informed consent ahead of time, not during the interview itself.
Conduct the interviews. Start by building rapport before diving into questions. Listen carefully. Stay open to where the conversation goes. Follow your guide, but explore relevant ideas that surface naturally.
Transcribe your interviews. Transcripts don't need to be perfect for initial analysis, but they do need to be usable. If you're unsure how to approach transcription, see our guide to getting research interviews transcribed.
Code and analyze the data. This is where patterns, themes, and meaning start to take shape. You'll read through interviews multiple times, apply codes to meaningful passages and compare patterns across transcripts. Delve always keeps your codebook and transcripts in view, so you can move between reading and coding without losing train of thought.
Present your findings. Translate your analysis into a narrative that clearly reflects participants' voices and addresses your research questions.
Qualitative analysis doesn't have to be overwhelming
Take Delve's free online course to learn how to find themes and patterns in your qualitative data. Get started here.
Semi-structured interview examples
These examples come from Lung and Liu's study on daily interactions between nursing assistants and elderly residents in nursing homes.
How do you perceive the daily interactions between you and nurse assistants/residents in the nursing home?
In what situations do you usually interact with each other?
How do you describe the relationship between you and the residents/nurse assistants?
How do you perceive the interaction that you have just shared?
The first question is the opening question, while others are part of the probing questions.
Where qualitative analysis software saves time
Semi-structured interviews generate rich data, but they also generate a lot of it. Those ten interviews can easily produce 100+ pages of transcripts. You need to go through every line, at least a few times.
Once transcripts start piling up, manual approaches like sticky notes, highlighters, Word documents or spreadsheets become hard to manage. You're scrolling through files trying to remember which interviews mentioned work-life balance, whether you've already coded this passage, and what you named that code about organizational culture three interviews ago.
This is where qualitative analysis software makes a real difference.
Code transcripts at the speed of thought
Delve is one of the easiest qualitative analysis tools to learn because it works like highlighting interview text. You just upload your transcript, highlight passages, and apply codes. You can then see where codes appear across all interviews instantly. Just click a code to see every passage you tagged with it.
For new researchers, this removes friction you might not foresee. There’s no complex setup or methodology jargon, and it’s web-based and accessible from any device. It’s easy (and free) to add peer debriefers. Your memos are easy to track from start to finish. All of this happens at your own pace.
If you’re doing collaborative qualitative analysis with a team, everyone sees the same project in real time. You can divide interviews among multiple coders, compare interpretations, automatically calculate intercoder agreement, or give your dissertation chair view-only access for feedback.
Don't have a peer debriefer available? Delve's AI can chat through your data, suggest codes, and apply your codebook across transcripts. Unlike tools like ChatGPT, it keeps all codes and transcripts in context as your analysis develops, so you stay focused on your actual data instead of drifting in generalities.
Try Delve free for 14 days
Your analysis will evolve as you understand your data better. As you merge redundant codes, create subcodes, and rename concepts, Delve works with you every step of the way.
Try Delve free for 14 days. You can upload an interview transcript and code a few passages in less than five minutes to see if it fits your research workflow.
No commitment. Cancel anytime.
References:
Dejonckheere, M; Vaughn, L. (2019). Semistructured interviewing in primary care research: a balance of relationship and rigour. BMJ Journals.
Lung, C; Liu, Justina. (2016). How the perspectives of nursing assistants and frail elderly residents on their daily interaction in nursing homes after their interaction: a qualitative study. BioMed Central.
Adams, W. (2015). Conducting semi-structured interviews. ResearchGate
Cite This Blog Article
Delve, Ho, L., & Limpaecher, A. (2022, April 6). Semi Structured Interviews https://delvetool.com/blog/semi-structured