The cognitive walkthrough is an analytical evaluation method in which experts step through a concrete user task one step at a time and check, for each step, whether a user would recognize the correct course of action, understand it, and receive feedback that it was successful. The focus is on learnability for users without prior experience.
Procedure and preparation
A cognitive walkthrough stands or falls with its preparation. Before the actual session, four things must be established: the user profile, including realistic assumptions about prior knowledge and experience; the specific tasks to be examined; the correct sequence of actions for each; and a sufficiently concrete design state: wireframes, a clickable prototype, or a device sample. The basis can also include requirements from the user interface specification, whose feasibility and comprehensibility are examined in the walkthrough. Compared with user studies, a cognitive walkthrough can be conducted very early, quickly, and with little effort, a major advantage especially in early development phases.
In the session, each action step is considered individually. The group consistently takes the perspective of a user who does not know the product and has not read the instructions for use. The classic focus is first-time use, but a walkthrough can in principle also be conducted for trained or experienced users. What is assessed is not whether a step works technically but whether such a user would find it on their own and interpret it correctly. At its core, this checks whether the user's mental model matches the logic of the system.
A small but mixed group has proven effective: a moderator who consistently asks the guiding questions, someone for documentation, clinical or application-side expertise, and representatives from development. The participation of development is valuable but carries a risk: Those responsible for the design tend to defend comprehensibility rather than examine it. The moderator must actively enforce this separation of roles.
The four guiding questions per action step
The core of the method is a fixed set of questions that is answered for each individual step. It maps the path from the goal of the action to the action performed:
- Is the user pursuing the right goal at all? Does what the user wants to achieve at this point match what the system provides as the next step?
- Does the user notice that the right action is available? Is the control visible, recognizable as operable, and not hidden or overlaid by other elements?
- Does the user associate the available action with their goal? Do the label, symbol, or position match what the user wants to achieve, in the user's own professional language and not only in the manufacturer's?
- After the action, does the user recognize that they have moved closer to the goal? Does the system give feedback that clearly confirms progress or clearly indicates an error?
If any of these questions is answered with no, a potential usability problem has been found, and it comes with a cause attribution already included. That is the real value of the method: It names not only that a step is difficult but at which stage between perception, comprehension, and feedback the problem arises, analogous to the stages of the perception-cognition-action model.
Analytical rather than empirical: where it fits
The cognitive walkthrough is one of the analytical methods: It is based on the reasoned assessment of experts, not on observed behavior of real users. This distinction matters from a regulatory perspective. Analytical methods yield hypotheses about usability problems; empirical methods yield evidence.
In practice, the sensible sequence is therefore clear: The walkthrough is used to eliminate obvious comprehension and findability problems before testing with real users. This substantially increases the yield of the subsequent formative usability evaluation. Test time with clinical professionals is too valuable to spend on problems that a structured expert analysis would have found beforehand.
Use in medical technology
For medical devices, the cognitive walkthrough is worthwhile above all where the consequences of a use error are serious. The critical tasks identified in the task analysis suggest themselves as a selection criterion for the tasks to be examined. They are already documented and justify the prioritization in a traceable way.
The method is particularly revealing for products with very different user groups. A walkthrough conducted once from the perspective of an experienced intensive care nurse and once from that of a layperson in home use regularly reveals that the same design is self-explanatory for one group and not interpretable for the other. Looking at exception paths is equally effective: Abort, resumption, alarm acknowledgment, and emergency use are rarely walked through systematically in regular reviews.
Strengths and limitations
The strengths lie in speed and accessibility: The walkthrough can be conducted without test participants, works on early drafts, can be repeated as often as needed, and, thanks to the fixed set of questions, delivers results that are easy to document.
The limitations are equally clear. The method predicts user behavior; it does not measure it, and the prediction is only as good as the assumptions about the users. It reflects neither time pressure nor stress, distraction, or use environment factors. It delivers no quantitative data. And it has a systematic blind spot for everything related to routine and repetition: The method is designed for learnability and says little about the error susceptibility of experienced users who operate a product daily.
Regulatory reference
IEC 62366-2 lists analytical methods such as the cognitive walkthrough as complementary methods within formative evaluation. For the formative phase according to IEC 62366-1, the use of analytical methods is possible and sensible; for the summative usability evaluation it is not.
The FDA Human Factors Guidance draws the same line: For human factors validation testing, it expects a test with representative users on the final design under realistic conditions. A cognitive walkthrough is therefore an instrument of the design phase and a contribution to the usability file, not proof of usability.
The cognitive walkthrough checks a concrete task step by step against four fixed guiding questions and thus specifically uncovers learnability problems, quickly, early, and without test participants. Its value lies in the fact that it names not only where things go wrong but at which stage between goal formation, findability, association, and feedback.
It is an instrument of the design phase, not a proof instrument: The summative evaluation remains empirical and with representative users.
Frequently asked questions (FAQ)
Does a cognitive walkthrough replace a usability test with real users?
No. It is an analytical method and provides reasoned assumptions about usability problems, not empirical evidence. The summative evaluation according to IEC 62366-1 is conducted with representative users on the final or production-equivalent product, and the FDA also expects a test with representative users on the final design for human factors validation testing. The walkthrough is the upstream step that makes such tests more effective.
How does it differ from a heuristic evaluation?
A heuristic evaluation checks a user interface as a whole against general design principles and covers a broad spectrum of deficiencies. The cognitive walkthrough, by contrast, is task-related: It follows a concrete sequence of actions step by step and answers the same set of questions for each step. It goes deep where the heuristic evaluation goes broad.
How many experts should take part?
A small group of three to five people has proven effective, provided they cover different perspectives: moderation, human factors, clinical or application-side expertise, and development. More important than the number is the consistent separation of roles: Those responsible for the design should not also judge its comprehensibility.
When is the right time in the development process?
As soon as an interaction concept is concrete enough that individual action steps can be named, that is, well before the first working sample. The earlier the walkthrough takes place, the cheaper changes are. A repeat after major design changes makes sense, as the method is easily reproducible.
Do you want to detect operating problems before expensive study time with clinical professionals begins? We make real use situations visible early and safeguard your design decisions.
More about our UX researchSources
- IEC 62366-1:2015+AMD1:2020, Medical devices, Part 1: Application of usability engineering to medical devices
- IEC/TR 62366-2:2016, Medical devices, Part 2: Guidance on the application of usability engineering to medical devices
- FDA Guidance: Applying Human Factors and Usability Engineering to Medical Devices