Development and Validation of the Collision Anxiety Questionnaire for VR Applications

Title of the Paper

Development and Validation of the Collision Anxiety Questionnaire for VR Applications

Paper Information

• Field of Study: Human-Computer Interaction and Virtual Reality User Experience
• Keywords: Virtual Reality, User Experience, Collision Anxiety, Fear, Discomfort, Assessment, HMD, Quantitative Research

Research Background and Problem

• Identified Issues or Challenges:

  • The high level of immersion provided by virtual reality (VR) head-mounted devices makes it difficult for users to perceive their physical surroundings, increasing the risk of collisions with obstacles or other individuals in the real world.
  • Even the anticipation of collision risks can negatively impact user experience and reduce confidence in VR systems.
  • Currently, there is a lack of measurement tools specifically designed to assess collision-related anxiety (collision anxiety, CA), which may significantly affect the quality of VR user experience.

• Importance of the Problem:

  • Collision anxiety may act as a "latent confounding variable" affecting VR user immersion, fluidity, and behavior, necessitating clear identification, precise measurement, and validation.
  • In multi-user VR environments or dynamic interactive scenarios (e.g., potential collisions with physical obstacles or other individuals), anxiety states may exacerbate these effects.

• Research Motivation and Related Work:

  • Existing studies have noted that users exhibit cautious movement, tension, or even fear in VR due to the presence of obstacles or other individuals. However, these studies primarily rely on qualitative interviews or general emotional scales not specifically tailored to VR, failing to comprehensively capture collision anxiety.
  • Ring and Masuch (2023) proposed an initial version of the collision anxiety questionnaire (alpha CAQ) as an innovative attempt, but it has not undergone formal academic validation.

Solution

• Proposed Solution:

  • Develop and validate the Collision Anxiety Questionnaire (CAQ) to quantify users' anxiety across different dimensions related to collisions in VR.
  • Through exploratory and confirmatory factor analysis, refine and validate the CAQ's three subscales:
    1. General Collision Anxiety: Addresses fear and behavioral restrictions related to physical obstacles.
    2. Orientation: Assesses users' ability to perceive their position relative to obstacles or boundaries in the physical room while in VR.
    3. Interpersonal Collision Anxiety: Focuses on anxiety related to potential collisions with other players or bystanders.

• Innovative Aspects:

  • The first structured measurement tool specifically targeting collision anxiety, with VR scenario-specific relevance to user experience.
  • Introduces a dimension of interpersonal interaction anxiety, distinct from traditional collision questionnaires.
  • Incorporates reverse-scored items based on feedback to capture users' "orientation" tendencies.

• Implementation Steps:

  1. Improvement of the Initial Questionnaire:
     • Structurally modify the 11 items of the alpha version, adjust wording, reduce the number of points on the Likert scale, and add items related to interpersonal collision anxiety.
  2. Preliminary Experiment and Exploratory Factor Analysis:
     • Sample: 159 participants combining online and laboratory-based game experiences; responses to the extended CAQ 1.0 items.
     • Statistical Methods: Principal Component Analysis and Exploratory Factor Analysis to eliminate ineffective items and form a new factor matrix (CAQ 2.0).
  3. Validation Experiment and Factor Confirmation:
     • Design a two-player cooperative VR escape game experiment to further confirm the measurement quality of revised questionnaire items.
     • Use internal consistency (Cronbach's α) and validity tests (e.g., AVE, HTMT ratios) to confirm questionnaire quality.

Research Results

• Main Findings:

  • The final CAQ consists of 10 items divided into General Collision Anxiety (4 items), Orientation (3 items), and Interpersonal Collision Anxiety (3 items).
  • Confirmatory factor analysis demonstrated a well-fitting model (CFI, RMSEA, and other indicators met recognized standards) with acceptable internal consistency and discriminant validity.

• Experimental and Evaluation Results:

  • Negative Correlation Between Collision Anxiety and VR Experience: Laboratory participants (with limited VR experience) reported significantly higher CAQ scores compared to online participants.
  • Importance of Orientation: Reverse-scored orientation items received broad support, indicating that lack of orientation is a major source of collision anxiety.
  • Weaker Interpersonal Collision Anxiety: Particularly in single-player VR scenarios, this factor had reduced impact; results suggest the need for larger samples and multi-user scenario studies for further validation.

• Limitations and Future Research Directions:

  • Small sample size (97 participants in the first experiment, 62 in the second), though the factor model remained stable. Larger-scale, cross-cultural validation is needed in the future.
  • Recommend exploring the relationship between CAQ and physiological indicators (e.g., skin conductance) as well as VR user performance for a more comprehensive assessment of collision anxiety's impact.
  • Extend the questionnaire to augmented reality (AR) application scenarios and develop multi-language versions for broader validation.