A cognitive walkthrough is only as useful as the record it leaves behind. As you step through each action in a task and ask the guiding questions — will the user know what to do, will they notice the control, will they understand the feedback — the moments where the answer is "no" or "probably not" are your failure points. This topic is about recognizing those moments, naming them precisely, and writing them down in a form that another designer can act on.
A failure point is more than "this screen is confusing." It is a specific breakdown at a specific step, tied to a cause: the user cannot form the right goal, cannot find the control, cannot connect the control to their intention, or cannot tell from the feedback that they made progress. Categorizing failures by type keeps your findings diagnostic rather than vague, and a consistent documentation template (step number, expected action, what fails, why, evidence, proposed fix) makes your results comparable across evaluators and reusable later in the design cycle.
The last piece is severity. Not every failure deserves the same attention, and a list of twenty undifferentiated problems gives a development team no guidance. As you watch the videos below, pay attention to the criteria used to rate severity, the language used to describe a problem without blaming the user, and the distinction between findings produced by inspection methods like the walkthrough and findings produced by testing with real users.
A short Nielsen Norman Group video on assigning severity ratings to design issues found in usability studies or heuristic evaluations. It explains the small set of criteria evaluators typically weigh — such as how many users are affected, how much the issue impedes task completion, and whether it is a one-time or persistent obstacle — and how those criteria combine into a priority ranking. The focus is on reporting findings so that teams know what to fix first.
This is the most direct match for the "Distinguishing Severity Levels" subtopic and for turning a raw list of failure points into a prioritized set of recommendations; NN/g is the standard reference source for usability evaluation practice.
A five-minute NN/g video covering common mistakes made when talking to users, including leading rather than probing, over-explaining, and filling silence in ways that distort what the participant tells you. It contrasts neutral probing questions with questions that plant an answer. The emphasis is on gathering evidence that reflects the user's actual experience rather than the researcher's assumptions.
The same discipline of neutral, non-leading language applies when you write failure descriptions — describing the observable breakdown instead of asserting your preferred explanation keeps your documentation credible and useful.
This video distinguishes user testing from usability testing, two terms that are frequently confused, and explains what question each method is designed to answer. It situates both within the broader UX research process and discusses when a team would choose one over the other.
Knowing what usability testing does and does not tell you clarifies the place of an inspection method like the cognitive walkthrough, and helps you label your findings honestly as predicted failure points rather than observed user behavior.
Together these videos reinforce the documentation and prioritization half of the module. The severity-rating video supports MO3 directly by giving you explicit criteria for ranking findings by user impact, and it supports MO2 by showing what a finished, reportable issue entry should contain. The interviewing video models the precise, non-leading language that makes failure descriptions written during a walkthrough defensible, again serving MO2. The user-testing-versus-usability-testing video frames where inspection methods such as the cognitive walkthrough sit among evaluation approaches, which supports MO1 by helping you recall why the method's guiding questions are structured as predictions about user reasoning rather than observations of users.
These videos were chosen from YouTube on October 3, 2026. Each has captions; the player turns them on. Videos on YouTube can be changed or removed — if one no longer plays, tell your instructor. If a player does not appear, use its Watch on YouTube link.