Additions by Usability Analysis & Design, Xerox Corporation
- When you are screening the participants for a study, notice how they respond to your questions. Decide on a strategy for engaging the participant before they arrive for the usability study.
- Be careful of the social dynamics you set up with the participant
Don’t joke, indulge in sarcasm, flirt, or betray your own nervousness
Maintain a professional, neutral persona
Keep yourself “small” in relationship to the participant. Sit slightly back from the participant, in a chair that is lower.
Avoid wearing heavy perfume or aftershave. The participant may have allergies to the odor or find it distracting.
Don’t wear suggestive, revealing, or tight, uncomfortable clothes.
- Don’t bias the participant
Don’t betray your own views or opinions of either the participant’s level of skill
Don’t let the participant become aware of any bias you may have about the product.
- Avoid interactions with the user that can shift the focus from the user’s domain to the designer’s
Don’t expect the user to tell you how to fix problems
Don’t expect the user to answer other design questions
Always keep the focus of attention on the user, not yourself. Avoid “I” statements and long explanations of how the system works.
Stay in the relationship with the participant. Don’t worry about the next question you are going to ask.
Write down design ideas so that you don’t need to worry about forgetting them after the test.
- Don’t let yourself get impatient!
When the participant seems to have a problem, they can often unravel it without your help.
When you feel you should jump in, count to ten first.
If you jump in too soon, you lose valuable data and they become dependent on your help.
- Learn to probe in a neutral way to get information on which to base your design improvements
Techniques that encourage thinking out loud
- “Conversational disequilibrium”
- Summarizing at key junctions
- Focus on tasks, not features
Don’t ask “Do you like that dialog box?” but “Did that dialog box help you reach your goal?”
- Focus on questions, not answers
- Explore user thinking in a neutral way
Don’t be too quick to assume that the user is lost or having a problem.
Don’t say, “What is your problem here?,” but ask, “What is your goal?” or “What are you thinking you should do here?”
- Don’t betray your own interests or point of view by your comments, emphasis, “waking up” and getting interested, showing in facial expression or vocal tones that you disagree
- Good user-focused questions:
What is your goal?
What did you expect when you did that?
How did you expect that to work?
Can you tell me what you were thinking?
What do you want to accomplish here?
Describe the steps you are going through here.
How did you feel about that process?
Tell me about your thinking here.
What did you expect to happen when you . . . .?
- Repeat their own word or phrase back to them as a question: “That message is confusing?”
Echoing sets up a social dialog and reinforces social conversation expectations: they say something, you repeat it, they say the next thing because that is what is expected in conversation.
- Don’t put words in their mouth, or offer interpretations
If they say, “I’m not sure what to do here,” don’t say, “So you are confused because the menu bar is unclear?”
If they say, “That didn’t happen like I expected, don’t ask, “So you thought that the task menu would be displayed here?”
- Signal that you re listening (Mmm hmm . . .)
- Let your statements trail off and end in an upswing, as if you were asking a question. The participant will usually complete your statement.
“And you were expecting. . .?”
“And your goal is . . .?”
- Signal that you are there, you are interested, but that it is still their turn to talk (mmm hmmmm)
- Speak softly
Summarizing at key junctions
- When you have learned something new that is key to understanding, summarize the event and the thinking that the user explored, very briefly. Users may offer more detail about their thought process.
- Keep the recorder on or keep taking notes after you think that the test session is finished. Users will often make interesting reflections about their processes during the casual remarks at the end of the session.