UXLX - 3: Usability testing bootcamp

The 2nd workshop I visited on Thursday was the "Usability testing bootcamp" by David Travis ..and here are my notes:

Usability is not JUST about making things easy.
- Just measuring Completion rate is not enough
- Just measuring Satisfaction is not enough
- Just measuring Time is not enough

Number of users for Usability Test
- Are 5 users sufficient? 5 users is based on the assumption that you develop iteratively.
- Steve Krug: "Testing one user early in the project is better than testing 50 near the end"

Phases of the usability test
- Identify the test goals
- Recruitment
- Identify the test tasks
- Preparation
- Consent: If the user is ok with the usability test
- Context-Setting: explanation of the method
- Demo Thinking aloud by the moderator
- Explicit Permission to start using the technology
- Practice thinking aloud with the participant (with an easy task that has nothing to do with the test itself)
- Task creation
- Moderated the test
- Survey
- Incentive

Usability test situations
- Portable test lab: Take the system to the user
- Single room setup
- Classic testing lab setup (one-way mirror)
- Classic benchmark test: Time on task - no think aloud
- Multi-room setup: Observers in another room
- Remote moderated test
- Remote unmoderated test

Writing test tasks: Headline Tasks ("red routes") are the most important things to test.
Ask the following questions...
- Is it really a red route?
- Is it specific and measureable? (is there a begin and endpoint)
- Does it describe a complete activity (integrated, not simple tasks)?
- Is the task "portable" to competitor products, systems or services?
- Does it include enough information to complete the task yet avoid hidden clues?

How to moderate a usability task
Hats of the moderator:
- Flight attendant: Welcome the participants, provide coffee, small talk,...
- Sports Commentator: Talk to inform the observers, Remind the user to think out loud, let the user to repeat the task in his own words, ...
- Scientist: Responsible for avoiding test bias and recording the data.
Br, Claudia