The first step in a User-Centered Design process is to empathize with the target audience of your system. Here, we introduce two qualitative methods to learn about users: observation and interview. Early in a design process, observation and interview are helpful to learn about user's characteristics (e.g., age, abilities), their goals/challenges, and the tasks they perform. Then, as the design process moves on, you can use these techniques to learn about your software system’s limitations (and good parts). These techniques are also called qualitative data gathering. You can perform interviews and observation either in a controlled environment (lab study) or field (field study).

Observation

In our daily lives, we do things by following both explicit instructions and employing practical knowledge that is not written anywhere. For example, when you are working on a course project as a team, you'd be nice to your teammates both because it is a socially accepted rule and because you know from your experience that discordant team relationships would deteriorate your team's productivity. Just like that, people carry out tasks by following both written rules and employing unwritten tacit knowledge.

Empathizing with users involves learning about both explicit and tacit knowledge. Explicit rules are easy to identify since these are well documented and/or widely accepted. But tacit knowledge is harder to unveil. The most effective way to discover this kind of information is to spend time with people and observe them to get to know what they don’t know how to tell you. Researcher Genevieve Bell calls this “deep hanging out.” See the following presentation by Ellen Isaacs explaining how researchers use observation to learn about tacit knowledge and challenges that people deal with in using existing technology:

https://youtu.be/nV0jY5VgymI

The technique described in the video is called ethnography, a form of observation (Preece et al., 2015, p.256). Around 2:50-6:25 of the video where she shows how people use a copying machine, we could observe and learn: "What goals are users trying to achieve?" "What are the tasks they are performing?" "What challenges or pain points does the design of the system (the copying machine) impose on the users?"

We are particularly interested in identifying recurring patterns — themes — in people’s behavior that we observe. Such recurring patterns can be processes, practices, slips, and mistakes:

• Process: A process is a set of well-known rules or a defined workflow. It reflects rules about how an individual or an organization operates. You could learn a process through observation or interviews, but you may be able to find it by reading documentation or textbooks. In the video above, an example of such a process is “following a written manual to copy and print a set of documents”.
• Practice: A practice is an undocumented steps or methods that people use to handle situations when processes fail. It is a tip or technique to handle exceptions and extemporize when the rules don’t work. By nature, it is not documented anywhere, so it’s hard to discover. But it’s often extremely important for a designer to understand and seek for those hacks that people use to get around limitations. Sometimes such unofficial knowledge manifests as post-it notes and unofficial notes that people use to support each other (e.g., “paper may be jammed! Open the tray and check the paper.”). We want to tease out such hidden tips to improve the design in the future.

User errors are an important source of information in thinking what to design and how to design. But it’s important to distinguish between two types of errors that people make: slips and mistakes.

• Slip: A slip is an error wherein a user knows how to accomplish a task, but something prevents them from completing it correctly. A slip occurs when a person intends to do one action but ends up doing something else. When you observe slips, you can often reduce errors by making small physical design improvements.
• Mistake: A mistake occurs when the wrong goal is established or the wrong plan is formed. In this kind of error, a user does not know how to accomplish a task correctly.

Conducting Interview and Observation

Interview is another method that is commonly used in collecting data to understand about users and their tasks. See the following video where a designer interviews (and observes) his participant to use a mobile application.

https://youtu.be/Qq3OiHQ-HCU

Here, the designer is combining an interview and observation. First, he asks an interviewee to answer simple demographic questions (e.g., "tell me what kind of work you do"). He then moves on and asks the interviewee to use a prototype application. Then he observes how she uses the application. Next, the researcher asked the users to speak out what she is thinking or feeling while interacting with the application in the video. The technique is called think-aloud, which lets you know what is happening in the person's head while using a prototype. Ask them if it is ok to record what they speak out so that you can record their thoughts and analyze them later.

https://youtu.be/jy-QGuWE7PQ

There are three interview types: structured, unstructured, and semi-structured.

• Structured interview: An interviewer only asks a predefined set of interview questions. It is like a verbal survey. Because it uses a predefined set of questions, it reduces the variability of the collected data due to interviewer’s experience, and thus increases the reliability and credibility of the data. However, its prevents the interviewer from drill down into something that might be interesting.
• Unstructured interview: An interviewer has a full control over what questions to ask during the interview. It increases the flexibility of the interview, but decreases the reliability of the data. It is used when it is hard to determine appropriate interview questions.