Remote and Collaborative Virtual Reality Experiments via Social VR Platforms

Title of the Paper

Remote Collaborative Virtual Reality (VR) Experiments via Social VR Platforms

Bibliographic Information

• Subject Area: Virtual Reality (VR), Human-Computer Interaction (HCI), Group Collaboration Research
• Keywords: Virtual Reality, Social VR, Crowdsourced Research, Replicability Studies, Fitts' Law, Collaborative Desktop Interaction, Quantitative Research, Qualitative Research

Research Background and Problem

• Problem or Challenge:

  • VR research is typically conducted in laboratories due to hardware requirements (e.g., head-mounted displays, HMDs) and the complexity of distributing applications, which limits sample size, collaboration, and access to experienced users.
  • The COVID-19 pandemic further restricted the feasibility of lab-based experiments.
  • Existing crowdsourcing methods primarily target traditional online users, with limited focus on users equipped with VR hardware.

• Significance:

  • VR user research is crucial for understanding user behavior and interaction with virtual technologies. Conducting VR experiments effectively in remote settings could overcome the aforementioned limitations.

• Motivation and Related Work:

  • Existing studies have explored remote user research using tools like Amazon Mechanical Turk, but most rely on smartphone-based VR devices.
  • Social VR platforms (e.g., VRChat) offer distributed synchronous virtual environments, large VR user communities, and user-generated content capabilities, potentially opening new avenues for research.

Solution

• Method and Approach:

  • The authors utilized the VRChat platform to replicate two classic experiments:
    1. Fitts' Law Extension: Research on target selection time in 3D VR environments.
    2. Collaborative Desktop Interaction: Study of interaction patterns and spatial arrangements among dyads solving network diagram problems.
  • Both studies validated the platform's effectiveness in supporting remote quantitative and qualitative research.

• Innovations:

  • Leveraging existing functionalities of social VR platforms (e.g., synchronous virtual environments, user-generated content) to design experiments, thereby overcoming challenges related to development and hardware distribution.
  • Pioneering remote experimental methods based on social VR platforms, opening new pathways for VR research.

• Implementation Steps and Techniques:

  • Technology and Platform Selection:
    • Used VRChat's Unity SDK to create experimental scenes and interaction logic while ensuring user privacy protection.
    • Controlled the experimental environment using synchronized variables and event handling mechanisms.
  • Experiment Design and Execution:
    • Quantitative research involved a 3D extension of Fitts' Law, designing 3D target selection tasks and collecting movement time data.
    • Qualitative research replicated a dyadic collaborative problem-solving task, analyzing user interaction behaviors and recording videos.
  • Data Collection and Analysis:
    • Data was collected through screenshots and in-game logs, manually extracted, and statistically analyzed to verify the success of experiment replication.

Research Outcomes

• Specific Results:

  • Both experiments (quantitative/qualitative) yielded results consistent with the original studies in remote settings, demonstrating the feasibility and effectiveness of conducting VR research via VRChat.
  • Confirmed that experienced VR users significantly optimized experimental performance, such as faster and more consistent completion times in the Fitts' Law experiment.
  • In collaborative tasks, participant behavior patterns (e.g., coupling styles, spatial arrangements) were similar to the original experiments but influenced by VR-specific characteristics (e.g., limited physical expression, remote nature).

• Advantages Compared to Traditional Methods:

  • Unlike traditional lab methods, no complex hardware distribution is required, enabling remote collaboration and real-time observation.
  • Direct access to a pool of experienced users with VR hardware lowers the barrier to participation.

• Experimental or Evaluation Results:

  • Fitts' Law Experiment: Successfully replicated Fitts' Law (ID as a significant predictor of movement time), with model accuracy reaching 77%-80%.
  • Collaborative Interaction Experiment: Replicated and validated the importance of "Same Problem-Same Area" (SPSA) collaboration from the original study, with findings adapted to VR environments.

• Limitations and Future Directions:

  • Limitations:
    • Data extraction relied on manual processes and has not been fully automated.
    • Platform limitations led to occasional data synchronization losses (approximately 1% of data).
    • Certain VRChat platform functionalities (e.g., HTTP request support) are not available, requiring temporary solutions from developers.
  • Future Directions:
    • Develop specialized social VR platforms for research or collaborate deeply with existing platforms to provide enhanced technical support and stability for experiments.
    • Explore collaborative interaction characteristics in VR environments, optimize experimental designs, and enhance user immersive experiences.