A monograph on semi-structured interview

As UX professionals, we are storytellers. Creating narrative through the medium of design in an effort to make interactions seamless. To weave such narratives, requires an in-depth understanding of the intended users. To gain dormant truths from users about issues, their motivations, frustrations, and intentions regarding a certain topic. Through all of my time conducting interviews, if there is one truth that I am certain of, is that of the importance of listening to your users. Getting said information out of users can be done in a myriad of ways, from ethnographic studies and naturalistic observation to casual conversations and interviews.

Interviews are generally categorized in 3 types 

• Unstructured conversation
• Semi-structured interviews
• Structured surveys

Structured interviews follow a rigid set of questionnaire that does not allow for mobility in questioning, conversely unstructured conversation can turn out to be so vague that often the focus of interview can be lost in vestigial conversation. Although all forms of interviews can be employed to suit appropriate instances, the focus of this monograph rests with Semi-structured interviews. A method of contextual inquiry that walks a fine line between providing structure and freedom to get the most pertinent information from users.

So then how to best conduct semi-structured interviews?

One of the pre-requisites in conducting semi-structured interviews is having a general idea of the kind of information that needs to be gathered. Based on that, an interviewer can prepare probe points that aim to explore necessary themes. The interviewer must judge areas of interests that the participant feels confident about, and then delve into those areas in detail. To an expert interviewer, response to a question leads to another pertinent question, akin to the branching of a tree. Each branch is a vein of idea to explore and gather insights from. A single participant might only be able to provide information on a single branch in great detail, or/and a number of branches with only a high level understanding of the problem, but if the sample size of the participants is big enough, those branches will come together to give the entire picture of the tree.

As UX professionals, we often get so involved in the intricacies of design, that we can sometimes forget that at the end of the day we are dealing with human problems and characteristics. As such, one of the better ways to extract insights out of human subjects is to make them feel comfortable. Therefore, it is of paramount importance to develop a rapport with your participants before or during the beginning of the session. An interviewer could start by making small talk about the weather, or inquiring with general personal questions to set an aura of comfort. An investment of time as little as 5 minutes can lead to a richer elicitation of information.

To supplement, an interviewer's most promising quality is being able to listen to what the participant is saying. The purpose of an interview is not to gather pieces of information but rather, gain true insight and a deeper understanding of what the participant is trying to say. Oftentimes, even though the users know what they want, they won't be able to elucidate it. It is the job of the interviewer to extract the meaning behind words. So, do the awkward pause and let the deafening silence make them uncomfortable.

Interviewers are human too, and as such they have a set of biases that define their reality and understanding of a concept. As UX professionals, it is essential that they not let their personal biases inform the interview process. Doing so otherwise would beat the entire purpose of placing end users at the center of our design process. Therefore, interviewers have to make a conscious and concerted effort to not ask leading questions during the interview. So instead of asking something like "Is this functionality good?", an appropriate question could be "What do you think of this functionality?".

Additionally, it is also advised that interviewer be mindful of time constraints. It is easy to get lost in conversation and dismiss the passage of time. Being aware of it helps keep focus on the important things that the interview aims to achieve. This is not to say that one should interrupt participants mid answer or force them to answer quickly, but rather just to be mindful of the clock ticking. Once the interview has been concluded, immediately jot down important learnings and then when time permits review findings. This will help in informing the probe points (if applicable) for the next session. If there is another session involving the same individual or group of participants , start by reiterating what happened previously. It's welcome revision and provides a buy in to start the current session.

Semi-structured interviews can be used during the Formative and the Generative stages of human centered design process.

Formative

During this stage of the design process, we aim to answer the following question, “How do I know something needs to be designed or could be improved?”. Semi-structured interviews can be employed to answer this question. A prerequisite for this process is to have a problem space in mind. Once we have that, we can then conduct interviews with experienced professionals in that particular problem space to know more about the problem. Exempli Gratia – Let us, for the sake of hypotheticals say that we are looking at the problem space of accessibility. Specifically, in the area of consumer electronics, say a smartphone or a wearable smart device.

A quick literature review would suggest that there is a need for devices with better accessibility, and in many cases, those devices have already been made. If one intends to improve those devices, then one can conduct semi-structured interviews with the original designers of the product to gather the thought process that went behind the making of such a product. Conversely, if no such solution exists in that particular domain, we can conduct semi-structured interviews with health professionals or individuals suffering from the said accessibility issue to gauge the extent of impact that such a solution could have. For instance, we can conduct semi-structured interviews with either individuals with cognitive impairments, or those who take care of them to understand whether or not a solution in that problem space is warranted.

Generative

From the example above, we see that semi-structured interviews can be most useful in determining the way we can deliver solutions. For cognitive impairments, we would be interacting with potential end users. They can tell us the best way in which a technology can help them improve their lives. This would involve

• Understanding the existing system of respite in place for such a problem.
• Finding out what methods are best suitable to improve upon those systems and if no such method exists, then through interviews ascertaining, what those potential systems could be.
• Figuring out the best way to deliver those solutions.

In conclusion

• Preparation is key.
• Pre plan a set of probe points (not questions per say) to elicit a more natural response.
• Develop a rapport with the participants to evoke a richer feedback.
• Be mindful of the time spent during the session. Breaks at the 30-40 min mark are most welcome.
• Post interview, review the findings, as it helps inform the probe points (if applicable) for the next session.
• Avoid leading questions
• During the beginning of a successive session, always reiterate what happened during the previous one. It’s welcome revision and provides a buy in to start the current session.