Reward Seeking or Loss Aversion?: Impact of Regulatory Focus Theory on Emotional Induction in Children and Their Behavior Towards a Social Robot

Title of the Paper

Reward Seeking or Loss Aversion? Impact of Regulatory Focus Theory on Emotional Induction in Children and Their Behavior Towards a Social Robot

Paper Information

• Research Area: Human-Computer Interaction and Emotional Education for Children
• Keywords: Social robots, human-robot interaction, emotional induction, regulatory focus, social engagement

Research Background and Problem

• Issues and Challenges:

  • Existing studies have demonstrated the importance of emotional induction in children's education and social development. However, how to apply the psychological Regulatory Focus Theory (RFT) to induce emotions during child-robot interaction (cHRI) remains unclear.
  • Early RFT research primarily focused on matching the robot's regulatory focus personality with the user, rather than exploring how to design interactions to induce emotions.

• Significance:

  • Positive emotions are closely linked to the development of children's social and cognitive skills, problem-solving abilities, creativity, and attention span.
  • Emotional induction has broad potential applications in children's education and social skills training.

• Research Motivation and Related Work:

  • Regulatory Focus Theory distinguishes between two behavioral motivations: promotion focus (seeking rewards) and prevention focus (avoiding losses).
  • Tasks designed with a promotion focus can enhance creativity and task performance (associated with positive emotions). However, how to integrate RFT and emotional induction into the domain of child-social robot interaction remains unclear.

Solution

• Method:

  • Designed two distinct interaction scenarios (promotion focus vs. prevention focus).
  • Used a "pretend play to complete tasks together" scenario, incorporating robot expressions and dialogue adaptation.
  • Extracted emotional and behavioral data from children's participation (facial recognition and behavioral coding).

• Innovations:

  • For the first time, interaction design was comprehensively based on RFT, rather than merely adapting the robot's personality.
  • Verified how RFT-based scenario design induces corresponding emotions in child-social robot interactions.
  • Expanded traditional emotional induction experiments (e.g., images and music) by introducing robot interaction as a new medium.

• Implementation Steps and Techniques:

  1. Emotional Induction Design:
     • Method: Combined "imagination" and "visual stimuli" to set up two interaction scenarios:
       • Promotion focus scenario: Rewards are given upon task completion.
       • Prevention focus scenario: Failure to complete the task leads to danger.
     • Emotional expressions: Displayed corresponding facial expressions and verbal behaviors through the EMYS robot.
  2. Task Flow:
     • Task introduction (pre-test storytelling).
     • Main task design (interactive storyline based on regulatory focus theory).
     • Post-task storytelling (post-test).
  3. Data Collection and Analysis:
     • Used Affectiva software to analyze children's facial emotion indicators (e.g., smiles, happiness, fear).
     • Behavioral coding tool ELAN recorded children's verbal behaviors (e.g., questions, responses, greetings).

Research Findings

• Specific Results:

  • Emotional Induction Study:
    • Successfully induced happiness in children during promotion focus tasks.
    • However, no significant fear induction effect was observed in prevention focus scenarios.
  • Social Engagement Study:
    • Children's social engagement was significantly higher in promotion focus scenarios compared to prevention focus scenarios.
    • Although induced happiness did not persist in post-task assessments, children exhibited more verbal engagement behaviors throughout the interaction.

• Advantages Over Existing Solutions:

  • Explored emotional induction through scenario design rather than solely focusing on robot personality, marking a new direction for RFT applications in HRI.
  • Utilized multimodal data, including facial expressions and verbal behaviors, providing robust evidence for the link between emotions and social engagement.

• Experiment and Evaluation Results:

  • Facial expression data indicated that children's happiness scores were significantly higher in promotion focus scenarios compared to prevention focus scenarios.
  • Voice behavior analysis showed that children interacted with the robot more frequently in promotion focus scenarios, aligning with successful emotional induction.

• Limitations and Future Directions:

  • Limitations in Fear Induction: Prevention focus did not significantly induce fear, possibly due to ineffective induction methods (e.g., scenario programming), facial recognition errors, or the absence of more sensitive measurement tools like EEG.
  • Lack of Emotional Retention: Induced happiness did not show lasting effects, suggesting future exploration of methods for prolonged emotional retention.
  • Future directions include analyzing children's storytelling creativity, enhancing the effectiveness of prevention focus induction, and evaluating the specific impacts of different induced emotions on task performance and social skill development.