IdeaBot: Investigating Social Facilitation in Human-Machine Team Creativity

Title of the Paper

IdeaBot: Investigating Social Facilitation in Human-Machine Team Creativity

Paper Information

• Subject Area: Human-Computer Interaction, Social Facilitation and Innovation in Team Collaboration
• Keywords: Chatbot, Team Collaboration, Social Facilitation, Creativity, Idea Generation

Research Background and Issues

• Challenges and Issues Identified:
  • There remain many unresolved mysteries in the study of human creativity, particularly in team collaboration, where both facilitative and inhibitory scenarios can occur.
  • Current research on the role of chatbots in team collaboration (especially in creative tasks) is limited, especially regarding their performance as active team members rather than passive assistants.
  • There is ongoing debate about whether AI robots should better mimic human interaction styles or retain their robotic characteristics.
• Research Importance:
  • As artificial intelligence expands into communication scenarios, understanding how robots influence the co-creation process between humans and machines can help design more effective collaborative systems.
• Research Motivation and Related Work:
  • Based on social facilitation theory and distraction-conflict theory, this study explores whether the presence of non-human teammates can alleviate social pressure and thereby enhance creative performance.
  • Investigates the potential impact of chatbots on team creativity through different conversational styles (robotic or human-like).

Solution

• Research Methods/Experimental Design:
  • The authors conducted three text-based chat experiments (cross-group studies) to explore factors influencing human-machine team collaboration:
    • Study 1: Compared collaboration outcomes with real chatbots and real human teammates.
    • Study 2: Engaged in human-like conversational styles with human assistants but informed participants that the teammate was either a robot (fake robot) or a human.
    • Study 3: Engaged in robotic conversational styles but informed participants that the teammate was either a human (fake human) or a robot.
• Innovative Aspects:
  • The unique contributions of this study include:
    1. Emphasizing the role of AI as an active participant in team collaboration, rather than a traditional passive collaborator or tool;
    2. Exploring the effects of "perceived identity" (viewing teammates as robots or humans) and conversational styles (human-like or robotic) in team tasks;
    3. Analyzing whether anxiety levels in group communication moderate the impact on task success and creativity.
• Key Techniques and Analytical Methods:
  • Collected and analyzed text chat records and participants' subjective feedback (e.g., creative self-efficacy).
  • Data analysis included independent sample t-tests, path analysis (using the R package lavaan), and hierarchical linear modeling.

Research Findings

Key Discoveries

1. Impact of Perceived Robot Identity on Creative Output:
   • Regardless of actual identity, participants who perceived their teammate as a robot produced higher creativity output in terms of quantity and quality (e.g., generating more original and informative ideas).
2. Moderating Role of Conversational Style:
   • When teammates adopted robotic conversational styles, the positive effects of social facilitation were stronger.
   • For participants with higher social anxiety, collaborating with robot teammates enhanced their sense of creative self-efficacy.
3. Perceptions of Interaction with Teammates:
   • In virtual experiments, robot teammates were perceived as more dominant, and this dominance was particularly effective in promoting team task performance when using robotic conversational styles.
4. Synthesis of Three Experiments:
   • Whether using robotic or human-like conversational styles, participants collaborating with robots consistently produced higher creativity output in terms of both quantity and quality compared to collaborating with human teammates.

Limitations and Future Directions

• Limitations:
  • The experiments were conducted via online text-based chats, which are limited to more formal and structured dialogue flows (lacking natural real-time interaction).
  • Did not compare results from face-to-face or other non-text communication methods.
  • The brainstorming tasks used for creative tasks were relatively simple and may not fully capture participants' divergent creativity.
• Future Research Directions:
  • Explore the impact of robots on more complex or unstructured creative tasks (e.g., creative writing, design tasks).
  • Enhance the ecological validity of experiments (e.g., using virtual reality or physical robots).
  • Investigate the long-term effects of robot teammate identity on continuous team collaboration behavior.

In conclusion, this study provides important insights for designing future collaborative chatbots, particularly in the selection of conversational styles and identity presentation methods. It also offers clear theoretical and practical guidance on leveraging robotic characteristics to enhance team creativity.