Research on Pediatric Social Robot Interaction: Rule-Based vs. LLM-Based Systems

Key Visuals

Background / Problem

Pediatric patients in hospital settings often experience significant psychological distress, and social robots have emerged as a promising tool for emotional support and intervention. However, most existing studies examine robot-mediated interactions at the group level, with limited attention to how individual child characteristics shape responses. This gap highlights the need for studies that incorporate standardized psychometric measures to better understand variability in pediatric robot interaction.

My Role

As part of a multidisciplinary research team at Samsung Medical Center, I was responsible for conducting the systematic literature review, administering usability and robot perception assessments during the experiment, and analyzing child responses from post-session video recordings. I am currently integrating these findings into a qualitative manuscript under preparation.

Process / Method

Individual characteristics were assessed using JTCI, CBCL, STAI, and a custom-developed pre-session child characteristics questionnaire. Usability and robot perception were measured via UEQ-S and Godspeed administered directly with child participants, and behavioral responses were systematically observed and coded from full-session video recordings.

Result / Outcome

The integrated analysis of psychometric profiles and observed behavioral responses forms the basis for describing individual difference patterns in children's reactions to robot-mediated interaction. These findings are currently being synthesized into a qualitative manuscript targeting submission to an SSCI-indexed journal.