Modeling Subjective User Experience in Haptics Through Individualized Bayesian Models

Area of Science:

• Human-Computer Interaction
• Robotics
• Haptics

Background:

• Traditional haptics evaluation prioritizes objective task performance.
• Subjective user experience (UX) aspects like emotion and preference are often secondary.
• A gap exists in unified frameworks for simultaneous evaluation of performance and subjective UX in haptics.

Purpose of the Study:

• To present a unified modeling framework for evaluating user experience (UX) in haptic and physical human-robot interaction.
• To simultaneously assess performance metrics, emotional response, and user preference.
• To model these metrics as a function of system design parameters.

Main Methods:

• Utilized Bayesian hierarchical modeling to capture individual variability and population trends.
• Conducted a line-following task with 30 participants using a custom haptic knob.
• Manipulated system parameters: damping, task difficulty, and performance feedback.

Main Results:

• Population-level models identified consistent effects on task performance.
• Emotion and preference showed significant individual variability, lacking consistent population trends.
• Individualized models revealed participant-specific patterns not evident in population summaries.

Conclusions:

• Unified modeling frameworks are essential for comprehensive haptic system evaluation.
• Individualized modeling is critical for capturing the full spectrum of user experience in haptics.
• Design and evaluation of haptic systems should account for substantial user-specific subjective responses.