Description
A short moderated segment in which participants narrate what a page or element is and predict what it does before clicking, exposing mismatches with their mental models.
At the start of a moderated session the participant looks at the home page or a comp without clicking and narrates whose site it is, what it is for and what they could do there. Krug fixes this at about three minutes and treats it as a realistic task, since every visitor must work out what a site is, and wrong first assumptions persist. Goodman and colleagues extend it into a guided tour: element by element and screen by screen the participant says what each control does, what a label means and what page would appear, all before acting. The rigour lies in capturing expectations before any action, because performing an action permanently alters the perception of its effect. Comparing predictions with the real behaviour yields a map of label and mental-model mismatches, plus a reading of the participant's domain knowledge for interpreting later task results.
Attributes
- Discipline
- Usability & Evaluation
- Loss type
- Active
- Stage
- Design, Evaluation
- Data
- Qualitative
- Effort
- Hours
Steps
- After the warm-up questions, open the home page or key screen from a bookmark and ask the participant what strikes them, whose it is, what they could do there and what it is for.
- Allow scrolling but no clicking; if they click, ask them to return, and stop after about three minutes (four at most).
- Optionally give one quick general task of about five minutes to see unprimed behaviour.
- Tour the main screens element by element for about ten minutes, asking what each item does in a couple of words, what labels mean and, without clicking, what page would appear and what it would contain.
- For each screen ask what matters most, what is missing, what they would do next and how they would get back to home or another major section.
- When an action is finally taken, ask whether the result was what they expected, then run the remaining tasks.
- Log every mismatch between prediction and actual behaviour, and the participant's apparent domain knowledge, for the debrief.
Sources
- Rocket Surgery Made Easy (Krug, 2010)
- Observing the User Experience (Goodman et al., 2012)
Also known as
- Home page tour
- Narrated first-impression walkthrough
- Hybrid interview
- Interface walkthrough probes
- Per-element expectation probes
Categories
- User Research
- Usability & Evaluation