Interviews
- Philip Northgraves
- Ned Potter
‘Interview’ is a rather formal term for what is essentially an in-depth conversation with a participant. There are three categories: a Structured Interview in which you follow a set list of questions, an Unstructured Interview in which you have no preset questions at all, and a Semi-Structured Interview in which the conversation is led by the participant, but if certain topics don’t come up naturally then you introduce them.
We would always recommend using Semi-Structured or Unstructured interviews as this is what ensures you truly get to hear about the user’s own experiences (rather than the conversation being driven by the researcher’s preconceptions or agenda).
Although it’s not essential to pair an Interview with a Cognitive Map, we would recommend doing so wherever possible. The conversations you have after a Cognitive Map has been drawn by the participant are markedly different from those which launch into questions straight away.
You will need to book a meeting room as Interviews usually happen in a one-on-one setting. They also require specific and informed, signed consent from the user. We update our Consent Forms regularly, so always go to this folder to find the most recent ones - make a copy and adapt the info for your project.
We would always recommend using interviews as part of UX projects as they’re often the key part of gaining an understanding of why users behave as they do, and what we could do to improve our services.
You don’t have to do a huge amount of work to learn a lot. For most projects 15 interviews with be enough to create an absolute wealth of useful information and insights you can act on. There is evidence to suggest you can learn 80% of what there is to learn about problems with a product or service from just 5 people. In the recent 'YorSearch Improvement Project' we used ethnographic techniques (Cognitive Maps followed by Semi-Structured Interviews) with 15 participants, but did so in three groups of five in order to get specific feedback at different stages of the project. This worked really well so is worth considering as a model to adopt for large scale projects.
You will need a number of materials to carry out Interviews:
Preferred way of taking notes | You will need a pen and paper or device to take notes with, and a device on which to record the audio of the interview |
| Audio recording device | A device on which to record the audio of the interview |
| Consent Forms | Two consent forms for the user to sign (one of which is theirs to keep). It’s good practice to send this out to the user in advance of the session, so they have time to review it. If using multiple techniques it is possible to roll the consent forms for each technique into one document. |
A session of Interviews typically includes the following steps:
Explain to the user what the wider project is and what you are trying to find out. Then explain the specifics of the Interview. Note that this step isn’t necessary if you’ve already done it as part of Cognitive Mapping exercise directly before.
Let the participant lead the direction of the conversation and talk about what they want to talk about (even if the relevance to your project isn’t immediately apparent). Take notes of things you want to come back to, and let them speak as much as possible.
The aim is to ask Open Questions as far as possible, and try and get more of a holistic overview rather than being very specific. For example you might ask a student ‘tell me about the process of completing an assignment’ rather than ‘what e-resources do you use’.
It’s good practise to conclude the interview by asking ‘is there anything else you’d like to talk about?’
As a side note it’s worth practising Interviews on a fellow member of staff. It can help get over the initial awkwardness of 1 to 1 conversations in a research setting, and crucially it allows you to work on asking open questions rather than leading questions. This doesn’t always come naturally, and is a skill to be worked on!
Think of a Master / Apprentice scenario. The user is the Master of their own lived experience of the Library, and you are the Apprentice seeking to learn how it all works for them.
It can be intimidating to analyse even a small amount of interviews because there is so much data. You can do it in basic way by just looking for themes and making notes on these as they emerge, but generally it’s worth trying to properly code the process the data - read the Qualitative Data Analysis page for a guide on how to do this.
The main variations are around how free-form the interview is going to be. As discussed above, we wouldn’t recommend Structured Interviews - the traditional format of having a set of questions and running through them in order. You risk missing out on really interesting insights if you lead the conversation in this way; the idea is to let the user lead.
An Unstructured Interview is completely free-form, and Semi-Structured Interviews involve having a few topics or questions written down and, if the conversation does not arrive at these naturally, asking them towards the end. We’ve most often used Semi-Structured, but Unstructured can be really effective when you don’t need to guarantee touching on certain topics.