Working Together Toward Interdependence: Chatbot-Based Support for Balanced Social Interactions Between Neurodivergent and Neurotypical Individuals

Research Background and Issues

• What problems or challenges did the authors identify?

  • Significant barriers exist in social interactions between neurodivergent individuals and neurotypical individuals, including differences in communication styles, social expectations, and uncertainty surrounding these interactions.
  • Existing social support technologies and interventions predominantly focus on helping neurodivergent individuals adapt to neurotypical social norms, neglecting the importance of mutual communication and understanding.
  • The "one-way adaptation" framework underlying these technologies may perceive neurodivergent behaviors as "deficiencies to be corrected," leading to psychological stress and potential social isolation.

• Why is this issue important?

  • Neurodivergent individuals are more likely to experience isolation and rejection in social settings, negatively impacting their quality of life and mental health.
  • Social interaction is inherently bidirectional, and inclusive, resilient communities can only be created through mutual effort and understanding.

• Research Motivation and Related Work

  • Researchers propose adopting the "double empathy" framework, emphasizing mutual adaptation rather than solely requiring neurodivergent individuals to conform to neurotypical behaviors.
  • The authors aim to explore a balanced design approach using chatbots to facilitate relationships between neurodivergent and neurotypical individuals, reducing uncertainty and differences in social expectations.

Solution

• What methods or solutions did the authors propose?

  • The authors designed three distinct social support chatbots (Friend, Coach, Personal Assistant) to address key barriers in social interactions.
  • A 5-day diary study and semi-structured interviews were conducted to understand the perspectives of neurodivergent and neurotypical participants regarding social barriers and their needs and preferences for social support chatbots.
  • Design recommendations were proposed to prevent chatbots from replacing real interpersonal relationships, providing one-sided support, or solely promoting neurotypical standards.

• What is innovative about this solution?

  • It breaks away from the traditional "one-way adaptation" framework in technology, emphasizing the role of neurodivergent individuals as providers of bidirectional social support, fostering "interdependence" in social interactions.
  • Chatbots are introduced as tools to facilitate mutual support and reflection, rather than merely correcting behaviors or replacing human interactions.

• What are the implementation steps and key technologies used?

  • Design of three chatbot prototypes:
    • Friend: Focused on emotional support, mimicking the role of a friend and listening to user reflections.
    • Coach: Provides general social advice.
    • Personal Assistant: Offers personalized, practical support and social guidance.
  • Research process:
    • A 5-day diary study involving 16 participants (8 neurotypical and 8 neurodivergent individuals).
    • Semi-structured interviews to gather reactions and opinions on the different chatbots.
  • Data analysis:
    • Thematic analysis using independent coding and cross-researcher coordination to ensure reliability.

Research Outcomes

• What specific outcomes were achieved?

  • Key factors for social success:
    • Both groups emphasized the importance of "giving and receiving support," but neurotypical participants focused more on "giving," while neurodivergent participants highlighted the importance of mutual, reciprocal support.
    • The importance of social networks was universally recognized, but neurodivergent individuals faced significant challenges in accessing social network support.
  • Major barriers to social interaction:
    • Social uncertainty: Both groups experienced uncertainty in expression, actions, and understanding.
    • Conflicts in social expectations: Neurotypical individuals prioritized implicit social norms, while neurodivergent individuals focused on basic social etiquette.
    • Gap in formal social skills training: Neurodivergent individuals commonly received social skills training, whereas neurotypical participants rarely did.
  • Preferences and use cases for chatbots:
    • Neurodivergent participants preferred the Friend chatbot, while neurotypical participants leaned towards Coach and Personal Assistant.
    • Participants believed chatbots could assist with activity suggestions, interaction reflection, and mood enhancement.

• How does it compare to existing solutions?

  • It shifts away from merely correcting neurodivergent behaviors, promoting mutual reflection and adaptation from both parties.
  • It integrates emotional reflection and behavioral guidance to serve bidirectional support.

• What were the experimental or evaluation results?

  • 13 participants reported that the reflective questions posed by the chatbots were highly beneficial for enhancing self-awareness and understanding others.
  • The proposed chatbots effectively reduced uncertainty in social interactions and improved the frequency and quality of interactions.

• Limitations and future directions:

  • The sample size was limited and only included neurodivergent individuals with intellectual disabilities, which may not represent the needs of all neurodivergent populations.
  • Predefined conversation scripts could feel rigid; future work could incorporate real-time interactive language models to improve flexibility.
  • The study focused primarily on delayed social support; future research could explore the potential of real-time support technologies.

Conclusion and Outlook

This study systematically explores the barriers and solutions for social interactions between neurodivergent and neurotypical individuals. By innovatively designing social support chatbots, it provides practical evidence for reducing social uncertainty, fostering mutual understanding, and promoting interdependence. Future work could expand the sample groups, incorporate advanced natural language processing technologies, and test the application of these new technologies in diverse social scenarios.