Doorways Do Not Always Cause Forgetting: Studying the Effect of Locomotion Technique and Doorway Visualization in Virtual Reality

Title of the Paper

Doorways Do Not Always Cause Forgetting: Studying the Effect of Locomotion Technique and Doorway Visualization in Virtual Reality

Paper Information

• Subject Area: Memory research in virtual reality, the impact of event boundaries on cognition and memory in human-computer interaction
• Keywords: Doorway effect, virtual reality, memory, environmental boundaries, event segmentation, walking, teleportation, object recognition

Research Background and Problem

• Problem and Challenges: The traditional "doorway effect" theory suggests that crossing environmental boundaries (e.g., passing through a doorway) negatively impacts memory. However, in virtual reality (VR), due to the non-realistic nature of environment design and operational constraints (e.g., the use of teleportation or other non-traditional navigation methods), it remains unclear whether different transitions and boundary designs still trigger the "doorway effect."
• Importance of the Research: Understanding the mechanisms by which event boundaries affect memory in highly immersive virtual environments can help optimize VR environment design, particularly in applications where users need to retain information (e.g., educational or training scenarios).
• Motivation and Related Work: Previous studies have validated the doorway effect in virtual environments displayed on flat screens, but research in highly immersive VR is scarce and shows conflicting results. Additionally, the specific impact of navigation methods (e.g., walking or teleportation) and boundary appearances on the doorway effect remains unclear.

Solution

• Methods and Approach: This study investigates the doorway effect through two experiments: the first examines whether navigation techniques (natural walking vs. teleportation) influence the doorway effect; the second explores the impact of different doorway visualizations (no door, standard door, sliding doors [transparent/opaque], and virtual portals) on the doorway effect.
• Innovations:
  • A systematic comparison of the doorway effect under two common VR navigation techniques: walking and teleportation.
  • Investigation of the potential impact of various non-realistic doorway designs on the doorway effect, expanding the scope of event boundary research.
  • Use of a highly immersive VR environment for the experiments, providing greater ecological validity.
• Implementation Steps:
  • Experiment 1: 40 participants were randomly assigned to either the walking or teleportation condition. They completed tasks involving crossing a doorway or remaining in the same room, with object recognition error rates and reaction times recorded.
  • Experiment 2: 20 participants sequentially experienced tasks involving five types of doorway visualizations (no door, standard door, transparent sliding door, opaque sliding door, virtual portal), with error rates and reaction times similarly recorded.
  • Descriptive statistics, repeated-measures ANOVA, and Bayesian regression models were used to analyze the results.

Research Findings

• Key Findings:
  • Crossing a doorway, regardless of whether participants walked or teleported, did not significantly affect object recognition error rates or reaction times.
  • Different doorway visualizations (standard door, transparent/opaque sliding doors, virtual portals, etc.) also had no significant impact on the doorway effect.
• Comparison with Existing Research:
  • Unlike the significant doorway effect observed in 2D screen-based studies, this study did not find similar effects in highly immersive VR.
  • The findings align with some other VR studies that failed to observe the doorway effect, suggesting that the triggers for the doorway effect may be more complex than previously thought.
• Limitations of the Experiments:
  • Potential confounding factors, such as participants' strong memory motivation preventing event segmentation, or the assumption that "the doorway itself is perceived as an event boundary," may not have been adequately validated in the study design.
  • The significantly higher movement time in the teleportation condition compared to walking may have introduced potential biases in the results.
• Future Research Directions:
  • Further exploration of whether the conditions for event model segmentation in VR differ mechanistically from those in real-world environments.
  • More precise task designs to distinguish the effects of associated objects and dissociated objects on memory formation.
  • Testing the impact of the doorway effect on user learning and memory in practical scenarios, such as VR educational games.

Conclusion

This study challenges the universality of the doorway effect in virtual reality and highlights the complex relationship between environmental boundary design, user behavior, and memory mechanisms. Although no significant doorway effect was observed, the research provides new insights for VR interaction design, contributing to improved user experiences and enhanced task performance.