One size fits all? Transferring social mindfulness measures to HRI

Area of Science:

• Psychology
• Human-Robot Interaction (HRI)
• Robotics

Background:

• Psychological measures are increasingly applied to Human-Robot Interaction (HRI).
• The Social Mindfulness Paradigm (SoMi) assesses social mindfulness towards robots using vignettes.
• Previous studies using SoMi suggested humans lack prosocial behavior towards robots, but these findings may be confounded by using human-human interaction (HHI) items and lacking real-life interaction.

Purpose of the Study:

• To investigate the validity of the SoMi task for assessing social mindfulness in HRI.
• To determine if real-life interaction with robots enhances prosocial behavior.
• To examine the impact of revised SoMi items and prolonged interaction on social mindfulness and anthropomorphism.

Main Methods:

• Study One: Participants interacted with a social robot (Cozmo) for three days, assessing social mindfulness and anthropomorphism before and after.
• Study Two: Participants completed a revised SoMi paradigm with items valued for both humans and robots.

Main Results:

• Prolonged interaction with Cozmo did not increase social mindfulness but enhanced perceived anthropomorphism.
• Revised SoMi items increased overall social mindfulness compared to previous studies, but not in the anthropomorphic condition.
• Real-life interaction did not inherently foster prosocial behavior towards robots.

Conclusions:

• Real-life interaction alone does not guarantee increased social mindfulness towards robots.
• Item selection and valuation are critical for valid psychological measures in HRI.
• The study highlights the need for careful validation of HHI measures for HRI and consideration of interaction context for external validity, challenging the Computers Are Social Actors (CASA) theory.