Developing questions for a semi-structured interview

This post came about after a conversation with a colleague and friend who was asking for some ideas on how to develop questions for a semi-structured interview. I provided some ideas and later that day wrote some more things down I thought were worth considering. This three-part series is what came out of that conversation.

There are many different types of interviews, and your approach will depend on your main research question and methodology. Types of interviews can range from structured, semi-structured, narrative and group (though there are others). Not to mention telephone, online and face-to-face. Different interviews will produce different outcomes, and for this reason, it is really important to have a clearly defined research question. The one main research question will determine your method, and any other questions should be sub-questions. It is worthwhile considering the pros and cons of different methods before settling on one, this is a good example for interviews.

In this post, I share some ways that help me think about developing questions for semi-structured interviews. I like to think of the interview schedule like a map for the interview. I see it as a guide to help me ensure I have asked everything I need to in order to help address my main research question.

The table method

When I’m developing questions for a semi-structured interview, I sometimes start with a table. It helps me check that I have all the interview questions needed in order to address the topic.

Main research question here.

Sub-question/topic 1	Sub-question/topic 2	Sub-question/topic 3	Other

You can see it has the main research question at the top like a heading, and each sub-topic/question in a columns. You can include other columns to suit your research. Add whatever helps you formulate the interview questions.

Write down all the questions you can think of and put them in the appropriate column. Don’t worry too much about the wording or order of them at this stage. Some people might prefer to write a list of questions first and then populate the table. This works just as well.

After filling in the questions in this table, you might notice:

• Gaps. You might only have one question for a particular sub-topic, so many want to add another.
• Too many questions: Check for any overlaps or doubling up of questions and decide which ones will work best for the interview and research.

A word on the types of questions

There are so many different types of questions, so I won’t go into full details here. I find the list that Bryman provides very thorough and I highly suggest you take a look (available online). I also wanted to provide four broad categories:

Technical – These are where you ask someone to define things, for example their role, a definition, what a key term means to them.

Reflective questions: Where you want someone to reflect on a topic of idea.

Personal: General personal questions: These are surface level questions such as day to day life, thoughts and opinion on things, and includes general introductory questions.

Deep personal question: Deeper questions ask a person to open up about their feelings or thoughts on something, such as a challenge they may be facing or may have faced in their life. You are essentially asking them to be reflexive. These are best left towards the end of an interview once you have established and build a level of trust and rapport.

Open and closed questions: I’ve been told that closed questions aren’t appropriate for a semi-structured interview, however I think they can be useful. I’ve used closed questions (sparingly) in interviews and have found they have been effective. Do be cautious about using these if the question is sensitive or is there is a stigma attached to it as it may be too confronting. Sometimes this will depend on the interviewee, sometimes on the topic.

You likely won’t use all the different types of questions in every situation. The type of question you ask may need tweaking depending on the interviewee.

The structure of the questions

After you have drafted a list of questions, order them. Try and leave personal and reflective ones towards the middle-end, after there has been more rapport established between the interviewer and interviewee.

Also consider the flow of the interview when trying to identify the order of questions. You want something that flows like a conversation. If you find that one question jumps dramatically from one topic to another, consider re-ordering the questions to try and make them flow better, or insert another one that might be useful for bridging from one topic to another.

Do the check

Once you have your list of questions, look at them and consider:

• Using appropriate language: Ideally, you want to avoid language that is too technical or academic. Keep the language accessible and easy to understand. You don’t want interviewees having to interpret the question or made to feel incompetent. You also don’t want to be seen as talking down to your participants, so getting the tone right is important.
• Phrasing. Check that the types of questions you are asking will elicit the information that you need. You might need a combination of different types of questions in order to fully cover the topic, and this may vary between participants.
• Biases. Our biases can creep into our research in many different ways, one is through our phrasing of a question. Avoid leading questions where assumptions and biases are in the question itself. I see this so often in surveys. An example: Why don’t you like travelling on public transport? A better question might be How do you feel about travelling on public transport? There may be times it is appropriate to ask something in a leading way, and this is determined by your research question.

A practice run

I suggest doing a trial mock-interviews with friends or colleagues to see how the interview flows (even if they know nothing about the topic). It helps you get a feel for:

• the flow of the interview
• identifying any gaps in questioning
• Determine if any questions are not working

I remember doing my first mock interview with a friend who was completely clueless about the topic. Even though his responses were not relevant, it showed me that one question did not flow at all. It stood out like a unicorn’s horn. I ended up adding a “bridging” question to help it flow into the conversation.

Not set in stone

The development of the interview schedule is an ongoing process. Keep in mind that an interview is primarily a conversation. Every interview will be different. It will be rare for you to ask the same questions, worded in the same way, and in the same order. That’s why it’s called a guide. It is there to guide you. This is another useful resource if you are looking at how to develop your questions and do the interviews.

In all my years of research, I have never had two interviews the same. In fact, I don’t even always refer to the interview schedule. Though I do look over it at the end to make sure I have covered all the topics I that needed discussion.

Interviews take preparation beyond a good interview guide. As a researcher, we play an active role in the interviewer/interviewee process of knowledge creation. I talk about this more in the second blog post in this series, where reflexivity becomes particularly important.