Human–robot conversation annotation

Thank you for taking part! You'll first go through a short tutorial (5 example clips), then watch … short conversation clips of a person talking with a wellness-coach robot and answer seven quick questions after each. Total time: about 10–15 minutes.

What we consider a "social norm break"

A social norm break = the robot did something that most people would consider socially inappropriate, rude, or out of place in this context — not just unusual or robot-like.

What you'll answer for each real clip

  1. Did the robot break a social norm in this clip? (yes / no)
  2. How confident are you in that read? (low / med / high)
  3. What in the clip did you base that on? (said / sound / visual)
  4. What was the main thing the robot was doing? (one of 10 categories)
  5. In your own words: what went wrong or well? (short text)
  6. What should the robot have done instead? (optional)
  7. If a person — not the robot — did the exact same thing, would you have seen it differently? (true / false)

How your work is saved

Answers save automatically after every clip. You are recorded only by an anonymous session ID; no personally identifying information is collected.

Progress
0 / 0
clip 0 / 0

Watch the full clip once, then answer the questions below.

Q1. Did the robot break a social norm in this clip?

Reminder: a norm break = something most people would consider socially inappropriate, rude, or out of place — not just unusual or robot-like.

Q2. How confident are you in that read?

Q3. What in the clip did you base that on? Pick any that apply.

Q4. What was the main thing the robot was doing in this moment? Pick the one that fits best.

Q5. What went wrong, or what went well, about how the robot behaved? One line.

Q6. If the robot's behavior was bad, what should it have done instead? (optional)

Q7. If a person — not the robot — did the exact same thing, would you have seen it differently?