Presence is the term used to describe the sense of "being in" a virtual environment. HITLab researchers believe establishing a good metric of presence would be useful for measuring the goodness of virtual environments.

subject in chair Our current goal is to develop materials and procedures for assessing vestibular contributions to spatial cognition. Specifically, we want to evaluate computer-generated animations as potential tools for studying self-orientation and self-motion perception. Does a non-verbal perceptual reporting procedure using animations improve assessment of spatial orientation? Are reports reliable? Do reports confirm expectations based on mechanical stimulus to vestibular apparatus?

In this experiment, subjects seated on chair rotated clockwise or counter-clockwise around yaw axis at 50 degrees per second. While rotating, subjects roll head from left to right or vise versa. "Cross-coupling" produces illusory motion.

Subjects report real and illusory self-motion by moving a doll head in which a Polhemus tracker is embedded so that motion of the virtual head displayed on a monitor corresponds to their own perceived head motion.

TRACKER --> VIRTUAL HEAD --> RECORDER

The"Real-time" virtual head motion is generated using WORLD UP (Sense 8) at 24 fps. The data is then saved for later analyses by a WORLD UP virtual VCR.

head diagram Two experiments were run with this procedure:

Experiment 1

Experiment 2

The results were as follows: Experiment 1, across trial reliability: 0.50, illusory motion reported.

Experiment 1

STIMULUS ILLUSORY PITCH RESPONSE - %
(Rotation Direction -
Head Roll Direction)
(Expected / Observed)
FORWARD REARWARD
CW-R / CCW-L 0
24%
100
76%
CCW-R / CW-L 100
38%
0
62%

For Experiment 2, across trial reliability: 0.94, illusory motion reported.

Experiment 2

STIMULUS ILLUSORY PITCH RESPONSE - %
(Rotation Direction -
Head Roll Direction)
(Expected / Observed)
FORWARD REARWARD
CW-R / CCW-L 0
41%
100
59%
CCW-R / CW-L 100
38%
0
62%

We found that a non-verbal perceptual reporting procedure using animations is reliable if limit set of stimuli. Also, perceptual reports do not correspond to expectations based on mechanical stimulus to vestibular apparatus. This disagrees with previous reports using cross-coupled angular accelerations and is analogous to previous reports of perceptual uncertainty during vertical oscillation.

For our next experiment, we plan to compare perceptual reports with eye movements.