Semi-Structured Interview | Definition, Guide & Examples

A semi-structured interview is a data collection method that relies on asking questions within a predetermined thematic framework. However, the questions are not set in order or in phrasing.

In research, semi-structured interviews are often qualitative in nature. They are generally used as an exploratory tool in marketing, social science, survey methodology, and other research fields.

They are also common in field research with many interviewers, giving everyone the same theoretical framework, but allowing them to investigate different facets of the research question.

What is a semi-structured interview?

Semi-structured interviews are a blend of structured and unstructured types of interviews.

• Unlike in an unstructured interview, the interviewer has an idea of what questions they will ask.
• Unlike in a structured interview, the phrasing and order of the questions is not set.

Semi-structured interviews are often open-ended, allowing for flexibility. Asking set questions in a set order allows for easy comparison between respondents, but it can be limiting. Having less structure can help you see patterns, while still allowing for comparisons between respondents.

When to use a semi-structured interview

Semi-structured interviews are best used when:

• You have prior interview experience. Spontaneous questions are deceptively challenging, and it’s easy to accidentally ask a leading question or make a participant uneasy.
• Your research question is exploratory in nature. Participant answers can guide future research questions and help you develop a more robust knowledge base for future research.

Just like in structured interviews, it is critical that you remain organised and develop a system for keeping track of participant responses. However, since the questions are less set than in a structured interview, the data collection and analysis become a bit more complex.

Differences between different types of interviews

Make sure to choose the type of interview that suits your research best. This table shows the most important differences between the four types.

Advantages of semi-structured interviews

Semi-structured interviews come with many advantages.

• Best of both worlds

  Semi-structured interviews are often considered ‘the best of both worlds’. Combining elements of structured and unstructured interviews gives semi-structured interviews the advantages of both: comparable, reliable data, and the flexibility to ask follow-up questions.

• No distractions

  The ability to design a thematic framework beforehand keeps both the interviewer and the participant on task, avoiding distractions while encouraging two-way communication.

• Detail and richness

  While similar methods-wise to structured interviews, questionnaires, and surveys, semi-structured interviews introduce more detail and richness due to their more open-ended nature. Participants can be asked to clarify, elaborate, or rephrase their answers if need be.

Disadvantages of semi-structured interviews

However, semi-structured interviews also have their downsides.

• Low validity

  The flexibility of semi-structured interviews can also lessen their validity. It can be challenging to compare responses between participants, depending how far the interviewer departed from the predetermined list of questions.

• High risk of bias

  The open-ended nature of semi-structured interviews can lead to the temptation to ask leading questions, biasing your responses. Conversely, your respondents may also seek to give you the answers they think you want to hear, leading to social desirability bias.

• Difficult to develop good semi-structured interview questions

  Semi-structured interviews can be difficult to conduct correctly due to their delicate balance of prior planning and spontaneous asides. Every participant is different in their willingness to share. It can be difficult to be both encouraging and unbiased.

Semi-structured interview questions

Since they are often open-ended in style, it can be challenging to write semi-structured interview questions that get you the information you’re looking for without biasing your responses. Here are a few tips:

• Define what areas or topics you will be focusing on prior to the interview. This will help you write a framework of questions that zero in on the information you seek.
• Write yourself a guide to refer to during the interview, so you stay focused. It can help to start with the simpler questions first, moving into the more complex ones after you have established a comfortable rapport.
• Be as clear and concise as possible, avoiding jargon and compound sentences.

How to conduct a semi-structured interview

Once you’ve determined that a semi-structured interview is the right fit for your research topic, you can proceed with the following steps.

Step 1: Set your goals and objectives

You can use guiding questions as you conceptualise your research question, such as:

• What are you trying to learn or achieve from a semi-structured interview?
• Why are you choosing a semi-structured interview as opposed to a different type of interview, or another research method?
• What questions or topics do you anticipate you’ll have follow-up questions about?

If you want to proceed with a semi-structured interview, you can start designing your questions.

Step 2: Design your questions

Try to stay simple and concise, and phrase your questions clearly. If your topic is sensitive or could cause an emotional response, be mindful of your word choices.

One of the most challenging parts of a semi-structured interview is knowing when to ask follow-up or spontaneous related questions. For this reason, having a guide to refer back to is critical. Hypothesising what other questions could arise from your participants’ answers may also be helpful.

Step 3: Assemble your participants

There are a few sampling methods you can use to recruit your interview participants, such as:

• Voluntary response sampling: For example, sending an email to a campus mailing list and sourcing participants from responses
• Stratified sampling of a particular characteristic trait of interest to your research, such as age, race, ethnicity, or gender identity
• Convenience sampling of those around you, such as other students at your university

Step 4: Decide on your medium

It’s important to determine ahead of time how you will be conducting your interview. You should decide whether you’ll be conducting it live or with a pen-and-paper format. If conducted in real time, you also need to decide whether in person, over the phone, or via videoconferencing is the best option for you.

Note that each of these methods has its own advantages and disadvantages:

• Pen-and-paper may be easier for you to organise and analyse, but you will receive more prepared answers, which may affect the reliability of your data.
• In-person interviews can lead to nervousness or interviewer effects, where the respondent feels pressured to respond in a manner they believe will please you or incentivise you to like them.
• Videoconferencing can feel awkward or stilted, which could affect your results.

Step 5: Conduct your interviews

As you conduct your interviews, keep environmental conditions as constant as you can to avoid research bias. Pay attention to your body language (e.g., nodding, raising eyebrows), and moderate your tone of voice.

Relatedly, one of the biggest challenges with semi-structured interviews is ensuring that your questions remain unbiased. This can be especially challenging with any spontaneous questions or unscripted follow-ups that you ask your participants.

How to analyse a semi-structured interview

After you’re finished conducting your interviews, it’s time to analyse your results. First, assign each of your participants a number or pseudonym for organisational purposes.

The next step in your analysis is to transcribe the audio or video recordings. You can then conduct a content or thematic analysis to determine your categories, looking for patterns of responses that stand out to you and test your hypotheses.

Transcribing interviews

Before you get started with transcription, decide whether to conduct verbatim transcription or intelligent verbatim transcription.

• If pauses, laughter, or filler words like ‘umm’ or ‘like’ affect your analysis and research conclusions, conduct verbatim transcription and include them.
• If not, you can conduct intelligent verbatim transcription, which excludes fillers, fixes any grammatical issues, and is usually easier to analyse.

Transcribing presents a great opportunity for you to cleanse your data. Here, you can identify and address any inconsistencies or questions that come up as you listen.

Your supervisor might ask you to add the transcriptions to the appendix of your paper.

Coding semi-structured interviews

Next, it’s time to conduct your thematic or content analysis. This often involves ‘coding’ words, patterns, or recurring responses, separating them into labels or categories for more robust analysis.

Due to the open-ended nature of many semi-structured interviews, you will most likely be conducting thematic analysis, rather than content analysis.

• You closely examine your data to identify common topics, ideas, or patterns. This can help you draw preliminary conclusions about your participants’ views, knowledge or experiences.

• After you have been through your responses a few times, you can collect the data into groups identified by their ‘code’. These codes give you a condensed overview of the main points and patterns identified by your data.

• Next, it’s time to organise these codes into themes. Themes are generally broader than codes, and you’ll often combine a few codes under one theme. After identifying your themes, make sure that these themes appropriately represent patterns in responses.

Analysing semi-structured interviews

Once you’re confident in your themes, you can take either an inductive or a deductive approach.

• An inductive approach is more open-ended, allowing your data to determine your themes.
• A deductive approach is the opposite. It involves investigating whether your data confirm preconceived themes or ideas.

Presenting your results (with example)

After your data analysis, the next step is to report your findings in a research paper.

• Your methodology section describes how you collected the data (in this case, describing your semi-structured interview process) and explains how you justify or conceptualise your analysis.
• Your discussion and results sections usually address each of your coded categories.
• You can then conclude with the main takeaways and avenues for further research.

Example of interview methodology for a research paper

Let’s say you are interested in vegan students on your campus. You have noticed that the number of vegan students seems to have increased since your first year, and you are curious what caused this shift.

You identify a few potential options based on literature:

• Perceptions about personal health or the perceived ‘healthiness’ of a vegan diet
• Concerns about animal welfare and the meat industry
• Increased climate awareness, especially in regards to animal products
• Availability of more vegan options, making the lifestyle change easier

Anecdotally, you hypothesise that students are more aware of the impact of animal products on the ongoing climate crisis, and this has influenced many to go vegan. However, you cannot rule out the possibility of the other options, such as the new vegan bar in the dining hall.

Since your topic is exploratory in nature and you have a lot of experience conducting interviews in your work-study role as a research assistant, you decide to conduct semi-structured interviews.

You have a friend who is a member of a campus club for vegans and vegetarians, so you send a message to the club to ask for volunteers. You also spend some time at the campus dining hall, approaching students at the vegan bar asking if they’d like to participate.

Here are some questions you could ask:

• Do you find vegan options on campus to be: excellent; good; fair; average; poor?
• How long have you been a vegan?
• What is the single biggest factor that led to your decision to become vegan?
  • Follow-up questions can probe the strength of this decision (i.e., was it overwhelmingly one reason, or more of a mix?).
• Do you think that more people should be vegan?
  • Why?

Depending on your participants’ answers to these questions, ask follow-ups as needed for clarification, further information, or elaboration.

After conducting your interviews and transcribing your data, you can then conduct thematic analysis, coding responses into different categories. Since you began your research with several theories about campus veganism that you found equally compelling, you would use the inductive approach.

Once you’ve identified themes and patterns from your data, you can draw inferences and conclusions. Your results section usually addresses each theme or pattern you found, describing each in turn, as well as how often you came across them in your analysis. Feel free to include lots of (properly anonymised) examples from the data as evidence, too.

Frequently asked questions about semi-structured interviews

When should you use a semi-structured interview?

A semi-structured interview is a blend of structured and unstructured types of interviews. Semi-structured interviews are best used when:

• You have prior interview experience. Spontaneous questions are deceptively challenging, and it’s easy to accidentally ask a leading question or make a participant uncomfortable.
• Your research question is exploratory in nature. Participant answers can guide future research questions and help you develop a more robust knowledge base for future research.

What are the four main types of interviews?

The four most common types of interviews are:

• Structured interviews: The questions are predetermined in both topic and order.
• Semi-structured interviews: A few questions are predetermined, but other questions aren’t planned.
• Unstructured interviews: None of the questions are predetermined.
• Focus group interviews: The questions are presented to a group instead of one individual.

What is an interviewer effect?

The interviewer effect is a type of bias that emerges when a characteristic of an interviewer (race, age, gender identity, etc.) influences the responses given by the interviewee.

There is a risk of an interviewer effect in all types of interviews, but it can be mitigated by writing really high-quality interview questions.