Are avatars treated like human obstacles during aperture crossing in virtual environments?

The current study set out to determine whether individuals walking in a virtual reality environment pass through apertures made of two avatars differently than apertures created by two pole obstacles, as previously observed between pole and human obstacles in real-world environments.

Eleven healthy young adults wore a head-mounted virtual reality display, walked along a 10 m path and passed through a virtual aperture located 5 m from the starting location. Participants were instructed to avoid colliding with the obstacles when passing through the aperture. The experiment was conducted in a block design, where the aperture was either created by two pole obstacles or by two avatars. In both conditions, the width of the aperture ranged between 1.0-1.8x each participant's shoulder width.

Regardless of whether the aperture was created by the virtual poles or the avatars, participants rotated their shoulders for all aperture sizes and results found no significant differences in shoulder rotation angle, onset of rotation, walking speed or velocity at time of crossing between the two types of obstacles. Therefore, it appears that the differences in avoidance behaviours observed in real-world settings between people and pole obstacles is not translated to a virtual reality environment.

It is possible that during experiments in which the avatars do not move, they do not possess human-like qualities suggested to be responsible for the increased caution used when walking through real human obstacles and instead, are treated as any ordinary obstacle.