How to do a semi-structured interview

Back to Library Back to Library

How to do a semi-structured interview

It’s possible to interview a stakeholder (a client, or someone deeply interested in your clients) and obtain information which can be used as data for reporting purposes. Done properly, this is called a semi-structured interview. Click here for important background information on semi-structured Interviews.

Here is a step by step guide to help you make sure that your interview produces relevant and good quality data.

Prepare for the interview

Why? The first step is to work out why you are doing the interview. It’s really important to keep the purpose of the interview in the front of your mind when you’re preparing questions, choosing interviewees, analysing the interviews and reporting on the information. This shapes your whole exercise.

Who? Once you know why you are doing the interview you can choose who to interview. The people you choose need a good understanding of the problem or issue you want to explore. They might be people using your service or people who have a concern about your clients.

How many? How many people are you going to interview?

Hint: You don’t need to interview lots of people to create a base of evidence. This is qualitative evidence, that is, it’s not about numbers. It’s about ‘rich descriptions’.

Here are some questions to help you decide how many people you’ll interview:

  • What other information is being collected? Are you adding to other information? If so, a few interviews might be enough. Is this the only information we are collecting? If so, you might need to do more interviews.
  • Are you getting information from the viewpoints of all your key stakeholder groups? If you aren’t, semi-structured interviews might be a useful way of filling any gaps.
  • What time and resources do you have to do this? Remember the 1 interview = 2 days’ work formula.
  • Once you have done some interviews, are you hearing the same information over and over? This is called ‘data saturation’. It means you can stop doing interviews because they aren’t giving you any new insights.

Once you have made these decisions, get the practical things organised. You need a quiet, safe and neutral place to do the interview, you might need to help people with costs such as transport or childcare, and you need a way of recording the interview.

Organise the ethics of the interview

You need to get informed consent. See How to get informed consent to gathering information which discusses issues like confidentiality, protecting the information you collect and the rights of the person you are interviewing.

Prepare the questions

To develop meaningful and relevant questions you need to have a good understanding of the topic you are researching. If you don’t have this from your own experience or reading, find some key people to talk to, and interview them informally to gain this understanding. If you are looking for someone external to your program or service to help prepare, or do, the interviews think about whether they have this level of understanding.

Prepare your questions so that they prompt people to give detailed answers. Try to avoid questions that might lead to ‘yes’ and ‘no’ responses. Examples might be:

  • What have you found helpful about the group?
  • What improvements would you like to see to our service?
  • Tell me about your experiences of working in partnership with our service
  • How has our program benefitted your clients?
  • What could have worked better?

Write your questions down. These questions are called your interview guide, or interview schedule. It’s important to test them before you start your formal interviews by running some 'mock interviews' with work colleagues, family members or friends. Testing your questions helps you to see which work and which don't and to refine them.

Once you have the structure prepared you can feel confident that you will be able to obtain the basic information you are looking for. You can feel free to ask the person you are interviewing extra questions to get them to expand on any interesting points.

Hint: Don’t ask 'double-barrelled' questions. These are questions which ask two things, eg. Tell me how you found out about our service and what made you decide to come? Ask these as two separate questions.

Hint: Don’t ask leading questions. These are questions that push someone to answer in a certain way, or imply that there is a right and a wrong answer, e.g. Did you find our service really useful? That must have been a very difficult time?

Do the interview

Spend a bit of time building rapport with your interviewee. Make sure you explain who you are and why you are doing the interview, how the information will be used and how you will manage the ethics of the process, such as obtaining informed consent from the person.

Ask the interviewee for permission to record the interview. Recording the interview can really help because you will miss things if you are trying to write down everything your interviewee says. Recording frees you up to listen, to think on your feet and to allow the interview to flow like a conversation. If the person you are interviewing doesn’t want to be recorded, ask if you can bring someone else into the interview to write the notes.

It can be useful to use prompts which include anonymous information from other interviews, for example, Other people I have interviewed have said that such and such was the case. Has that been your experience?. This can give people confidence to express their own views.

Don’t let the interview go for more than 90 minutes. People get tired. Make sure you leave time at the end of the interview to ask if there is anything your interviewee would like to add. It can help to summarise what you have been told as a way of reminding you both what has been covered.

Analyse and report on the information

This might sound difficult but it’s not if you take it step by step. Click here for How to analyse and report on qualitative information (stories, interviews or focus groups).

Other ways of using your information

You can also use your interview as one source of information for a case study. See How to write a case study.

You can write it up as a report. To do this, think about what story your interview has revealed. It might be a specific finding or it might be a pattern of findings.

You can’t turn your findings into statistics. For example, you can’t interview 10 people and then say ‘50% of participants said that transport was a problem’. You can say ‘participants interviewed reported that transport was a problem’.

You can use direct quotes from your interviews. You can show they are direct quotes by formatting them differently, for example:

               We were down the street one day and we ran into the NILS lady and she said that if we went to the Neighbourhood House they would be able to help us.

If you are writing your interviews up for a research paper or a journal article you will have to meet their formatting and style guidelines.

Hint: You will almost certainly have to make tough decisions and cut out details you think are extremely interesting. Not everything will be relevant to your key findings.

Other resources about interviewing

Harrell M & Bradley M, Data Collection Methods: Semi-structured Interviews and Focus Groups 2009, the Rand National Defense Research Institute.

Cohen D, Crabtree B. "Qualitative Research Guidelines Project." July 2006, Robert Wood Johnson Foundation.

Keller S & Conradin K 2010, Semi-Structured Interviews.

Laforest J 2009, Safety Diagnosis Tool Kit for Local Communities: Guide to Organising Semi-Structured Interviews with Key informants, Institut national de sante publique du Quebec.