Understanding how people react to alerts of incorrect movement, and how they respond to different biofeedback designs. 10+ rounds of user tests, successfully motivating 80% of users to click "Learn More".
Iteration 1: AR Biofeedback
Hypothesis: If the system informs users of what they should do to fix incorrect movements, users will be compelled to correct their movements
Key Learnings
Most participants find the side-view silhouette indication of “correct posture” confusing (doesn’t always fit bodyshape and disorienting)
Participants wanted immediate feedback about the details of their mistakes, as well as how to fix them
Static feedback (tick, cross marks) are ineffective in signaling the quality of users’ workout posture
Our hypothesis is validated that when people are told they are wrong, they are more motivated to take action and learn more about their wrong postures (80% + clicked "Learn More")
Hypothesis: If the system informs users of what they should do to fix incorrect movements, users will be compelled to correct their movements
Key Learnings
Hypothesis is initially validated that users will feel compelled when exposed to bodyless self-correct UI for asymmetry and audio instruction on how to correct posture
User reports cognitive overload and feeling distracted when attempting to fix her posture during one rep, but then being informed of another asymmetry in the next rep
User wants to see a consistent trend of asymmetry at the end of a group of workouts (terminal feedback)
Possible error bound indication to account for any inevitable inaccuracy
Next Steps
We plan to keep refining the design of bodyless self-correct UI system and conduct more user tests.
