DesignqualitativeIntermediate

Expert evaluation: Cognitive Walkthrough

Evaluate learning ease of an interface for new users.

What it is

Main objective

Evaluate learning ease of an interface for new users.

Detailed description

Cognitive Walkthrough is an expert evaluation methodology that simulates the mental process of a new user attempting to complete specific tasks in an interface. This technique examines each required action from the perspective of cognitive theory, evaluating whether the user can predict the correct action, find the necessary controls, and understand system feedback. Research demonstrates its effectiveness in identifying learnability and discoverability problems (Interaction Design Foundation). It is especially valuable for evaluating complex flows and onboarding of new users.

When it fits

Early stages with low-fidelity prototypes.

•Early prototypes
•Learnability evaluation
•Interfaces for novices
•Educational products

Effort level

Medium

Recommended number of users

1–3 experts

What you gain and what you give up

Advantages

Doesn't require real users: It eliminates the difficulty and cost of recruiting participants, which allows it to be run at any time.
Early evaluation: It can be applied to specifications or mockups before any code is written, saving development costs.
Focus on learning: It is excellent for evaluating systems that users must learn to use on their own (exploration), with no manuals or prior training.
Identifying cognitive problems: It helps you understand why a user would fail, not just where.

Disadvantages

Doesn't reveal every problem: Being a simulation, it does not capture the unpredictability of real user behavior or every possible usability problem.
Expert bias: Evaluators (experts) often find it hard to put themselves in the user's shoes, and may "hallucinate" problems that don't exist or overlook obvious ones (the designer's error).
Tedious: The process of analyzing each step with multiple questions can be long and tedious if the task is complex.
Limited focus: Originally designed for interfaces learned through exploration, it can be less effective for evaluating systems that require intensive expert training, though it remains useful.