Robotics

Synthetic emotions beat points for kids, but not for adults

New study reveals age shapes how social robots engage users best.

Deep Dive

A new arXiv preprint (2605.27539) by Morten Roed Frederiksen and Kasper Støy explores how to keep users engaged with small social robots designed to support children with anxiety disorders. The team built a tactile robot that uses either synthetic emotional feedback (e.g., happy/sad expressions) or gamified point rewards to encourage interaction. In a first study with 16 school children aged 6–8, participants showed a clear preference for emotional engagement over points. But that's only half the story.

In a second study with 14 university students aged 20–27 observed across a full day of interactions, the results flipped: points-based gamification produced significantly higher task accuracy (p<0.05) and better sustained performance over time. This contrast underscores a key insight for designers of socially assistive robots: stated preferences and actual behavioral outcomes can diverge depending on the user's age and interaction context. The authors emphasize the importance of validating engagement strategies through observed, naturalistic interactions rather than relying solely on self-reports, especially when building robots for therapeutic or educational use.

Key Points
  • Children (6–8) preferred emotional feedback over point rewards in a tactile social robot study.
  • University students (20–27) showed significantly higher task accuracy (p<0.05) with points-based gamification.
  • The study highlights that user preferences and actual performance can differ by age group, requiring context-specific design.

Why It Matters

Age-tailored engagement strategies are critical for social robots in therapy—one approach won't work for all users.