Understanding the basics of user research
Understanding the basics of user research
How to gather user data, validate it, and organize it.
As digital products age, they tend to compound in complexity. The rapid development of technology creates possibilities for products to combine features from previously separate products. Added features lead to added complexity, which in many cases, leads to products that are harder to use and can cause more errors. Product changes can even make the original features of the product perform worse. Despite good intentions, we are left with a bad user experience. Usability issues also arise because we tend to assume the users are similar to ourselves. For this reason, user research is crucial to gathering data about the users and the true driver of human-centered design.
User research is not just one method but a collection of many being both quantitative and qualitative. These methods vary from research conducted in the field to research conducted in laboratories, with the former being more common in product development. The goal is the same: knowing the users and how they are using the product.
In this article, I have summarized some of the methods you can to do successful user research. Combining different techniques and measures in an approach called triangulation is also covered, as well as ways of utilizing the gathered user research, giving you a sizable toolbox for doing user research in the wild.
From Contextual Inquiry to Usability Roundtables
We tend to link the word research to something that is done in a laboratory. This is the case in many sciences that aim to produce generalizable, reliable, and valid results needed to strengthen theories further. Most often, these laboratory studies are also highly controlled, and the amount of variable is reduced to a minimum. Although this approach suits the needs of basic sciences very well, it is not the best one for user research.
In user research, real-world context matters more than generalizable results. as users’ interactions often vary depending on the context. Many user research methods are therefore ethnographic by nature. The researcher observes society from the point of view of the subject of the study, which could mean observing the user using the product. Because context matters, the observation has to be done in the field. However, this observation in the wild comes with significant increases in time frames and higher costs. Contextual Inquiry | Usability Body of Knowledge *Contextual inquiry is a semi-structured interview method to obtain information about the context of use, where users…*www.usabilitybok.org
The answer to time and cost problems came in the form of Contextual inquiry, developed by computer scientist Karen Holtzblatt in 1986. As the name suggests, the main emphasis is in the context, as contextual inquiry provides the researcher with methods for extracting data about the user in context and more holistically. The approach aims to tackle the problem of users not being able to articulate detailed information on how they work. In Contextual inquiry, the researcher observes the user at work while the subject uses the product. The researcher is allowed to interrupt the user and ask questions concerning the use such as:
*“Why did you do it like that?”
- *“Was that expected?”
- “What is the next step?”
This way, the observer can get information about the product which was previously implicit to the user. The Contextual Inquiry session should take about two to three hours. The inquiry should be conducted by the product’s designers with Holzblatt and Beyer providing the following reasoning:
Contextual Inquiry has its problems as well. Users might for example consider Contextual Inquiry too invasive, not allowing the researcher to observe them work. Sending developers and designers out in the field may also turn out to be cumbersome and difficult to sell in some organizations. Concerns might also arise from privacy issues, scheduling problems, and time consumption.
There is a possible workaround to these problems, a method called** usability roundtables.** first conducted in usability laboratories, which was later changed to a more welcoming and pleasant conference room environment as the former was redeemed as too constraining.
The basic setting of a usability roundtable is that the user is asked to bring some examples from everyday work. Anybody from the design team wanting to ask questions from the user was then able to gather around the table and do so. This kind of approach is more like a meeting than testing. According to Butler, this made users more willing to participate in the design process, and the method was less time and resource-demanding than, for example, Contextual inquiry and results could be gathered more efficiently from specific issues.
There are limitations to usability roundtables as well. The approach requires a skillful operator to maximize the session’s benefit and to discourage team members from offering fixes for the user during the session. In addition to this, the users need to be selected so that they’re able to explain their work clearly, which could result in a bias towards more experienced users. The biggest problem, however, lies in the loss of context. Because the setting is different than where the user usually uses the product, some information may be lost.
Involving the User
Interviews, observation, and methods like Contextual inquiry and usability roundtables may help uncover issues in usability and introduce possible future additions to the design of the product. They, however, fail to truly grasp some of the users’ characteristics such as creativity.
Design Probes
Workshops and other similar means of involving the user can provide answers to this and are in heavy use these days. One such method is design probes. Defining what design probes are is rather hard as there are very few restrictions on what counts as a design probe. The main focus of these probes is to work as tools for design and understanding. As such, they can be objects, pieces of material, or pictures relating to the design at hand in some way. They pose questions that inspire the user to answer and provide solutions the designers can use as an aid. You can read more about designing probes here.
User Research With Drawings
Similars method of involving users in the user research process are acting and drawing:
Drawings can reveal emotional aspects, which could not be covered in word-based communication. This observation is especially effective when the emotional aspects of the product or service need to be studied, as these are also part of the user experience.
Both design probes and drawings have their shortcomings, which rise from the heavy emphasis on interpretation. Responses received from both vary according to personal and situational characteristics and may be hard to control. Researcher’s own interpretations, especially from drawings, may also create differing results.
Combining Research Methods
Data from one research tool will not most likely reveal all the aspects of users and their issues with the product. Using a single research method can also paint the product in a better light, or cause ***“inappropriate certainty” ***as Chauncey E. Wilson calls it his paper on his Triangulation. According to Wilson Triangulation means using multiple measures, methods, and approaches in the research process. Wilson lists different kinds of triangulation methods:
Research Method Triangulation: Using different research methods like questionnaires, focus groups, informal testing, and event logging.
**Facilitator Triangulation: **Different facilitators or evaluators with the same research method.
**Observer Triangulation: **Using different observers to record data from a given method.
**User Group Triangulation: ** Using multiple user groups with a single method.
Geographic Triangulation: Using participants from different locations.
**Qualitative-Quantitative Triangulation: ** Combining qualitative and quantitative approaches
There are often many sources of data available that can be triangulated right from the start. We can use bug reports, user group discussions, and articles combining them with findings from Contextual inquiry to find points where data converge. It is important also not to favor any method over another. Instead, when doing user research we should combine different methods with different strengths and biases and look for patterns.
Utilizing the data as a research tool — models and personas
Even if user research is done with multiple tools and by utilizing triangulation methods the data is nothing without if we don’t put it to use. The data has to be converted to a form capable of being communicated between the members of the product development team.
One such method for utilizing gathered user research data is affinity diagramming where the design team collectively writes all the points and issues, which have been noticed during the research process on post-it notes. The notes are then organized and grouped by the team. This allows even massive amounts of data to be utilized efficiently in the product development process.
Affinity diagramming only organizes data and presents it in detail. In order to better understand the user’s work structure, there are different means of modeling the users. These models are constructed from the data gathered about the users: Flow models (i.e. user flows) are used to inspect the models the users take and represent different types of users, while context models show how the organizations affect the user’s work. Physical models are used to represent the impact of the physical surroundings to work. Abstract work models are instead used to show relations between different aspects of work and sequence models communicate the steps needed for complementation of the task. Through modeling the user’s actions the team should be able to grasp the nature of the users’ work.
Another common method often used as a way to represent data is user personas, first coined by Alan Cooper in his book The Inmates Are Running the Asylum. These personas are constructed from research data and real stories gathered from users and possible from data collected through market surveys on similar large-sample segmentations studies. They are fictional characters with names and backgrounds used to represent users. The amount of information and how deep the personas go depend significantly on the design team’s requirements from the persona.
Others have argued against highly detailed personas. In the 2013 paper by William Hudson, Hudson argues against the idea of bloated a persona. Instead of placing weight on creating models of users, which easier to communicate, faster to develop, and better able to create a focus for the developers. Minimal collaborative personas should be short in the description and collaboratively created by the core team. This approach should allow a better sense of empathy from the team towards the personas with the help of persona stories instead of user stories. These persona stories describe common tasks, and goals users want to achieve. Hudson believes this will result in scenarios and visual designs that are more descriptive than prescriptive.
User research is vital for the user-centered design process Although it is not exact by nature, it still provides methods for constructing aids for the product development process. The user research methods qualitative nature, which, if used correctly, will reveal real information about the users, which quantitative methods could not. User research works best if the whole development team is part of it, as this may help uncover surprising points. For best results, users should also be employed through the entire development process and re-employed every time new the product is to be improved.
