Psychological evaluation of a group communication activation robot in a party game

Yoichi Matsuyama, Shinya Fujie, Hikaru Taniyama, Tetsunori KobayashiInterspeech 2010

TL;DR

This paper explores how a robot can actively participate in a group quiz game to enhance communication and engagement. They found that a robot's presence made participants feel happier and more "noisy" (in a good way!), suggesting a potential for robots to act as social catalysts in group settings.

What Problem Does This Paper Address?

Most current research on human-robot interaction focuses on one-on-one scenarios. However, a significant portion of human communication happens in groups. This paper tackles the challenge of designing and evaluating robots that can effectively participate in and even activate group communication, moving beyond simple task execution to fostering a more dynamic and engaging social environment. They specifically look at how a robot can act as a "communication activator" in a party game setting.

Methods

The researchers developed a communication activation system for a robot named "SCHEMA" to participate in a multi-modal conversation quiz game called "NANDOKU quiz." The system operates under a hierarchical constraint structure, ensuring the robot adheres to both game rules and group communication principles.

To evaluate the robot's impact, they conducted a laboratory experiment with human participants playing the NANDOKU quiz game in two conditions: one with the robot as a panelist and another without. They employed two evaluation methods:

  • Video Analysis: Participants' smiling duration was recorded and analyzed to objectively measure engagement.
  • SD (Semantic Differential) Method: Participants completed questionnaires using adjective pairs to subjectively rate their impressions of the interaction. They also provided free-form comments.

Datasets

Not discussed in the paper. The experiment used human participants playing a custom quiz game.

Key Insights and Ideas

  • Robots as Social Catalysts: The core idea is that robots can go beyond being passive observers or task-doers and actively contribute to the social dynamics of a group, making interactions more lively.
  • The "Noisy" Goodness: Participants reported feeling "more noisy" with the robot. This isn't about literal noise, but rather a sense of increased chatter, laughter, and general liveliness – a positive indicator of activated communication.
  • Pleasure Principle: The robot's presence significantly increased participants' feelings of pleasure during the game.
  • The Double-Edged Sword of "Ease": While the robot boosted pleasure and liveliness, it also decreased the perceived "ease" of the interaction for some. This suggests a trade-off that needs careful consideration.
  • Importance of Timing and Constraints: The authors identified that violations of group communication timing and constraints by the robot could negatively impact the interaction, leading to a decrease in perceived ease.

Key Results

  • Video Analysis: Smiling frequency was higher when the robot participated (13.2%) compared to when it didn't (11.1%), a statistically significant difference.
  • SD Analysis:
    • Participants reported feeling significantly more "pleased" with the robot present, particularly when one specific experimenter (Experimenter A) acted as MC.
  • Participants reported feeling significantly "noisier" (less silent) with the robot, indicating increased interaction.
  • The "ease" factor showed mixed results, with a slight decrease overall but maintained high levels with a more experienced MC (Experimenter B).
  • Free-Form Comments: Supported the quantitative findings, with participants noting increased liveliness and reduced silence due to the robot, but also mentioning issues with robot utterance repetition and timing.

Limitations and Future Work

The authors acknowledge that the robot's effectiveness in enhancing "ease" was limited, potentially due to violations of group communication constraints, particularly regarding timing and repetitive utterances. Future work will focus on analyzing the timing of robot behaviors in different situations to better understand and mitigate these issues, aiming to further refine the robot's ability to activate group communication without hindering the natural flow of interaction.

Summary updated

BibTeX
@inproceedings{matsuyama2010psychological,
  author = {Yoichi Matsuyama and Shinya Fujie and Hikaru Taniyama and Tetsunori Kobayashi},
  title = {{Psychological evaluation of a group communication activation robot in a party game}},
  booktitle = {Interspeech},
  year = {2010},
}
Figure