A primary advantage of the visual-inertial nulling procedure developed in Chapter 4 is that it factors out the strong inertial cues from gravity. However, this visual-inertial nulling procedure has serious drawbacks: it requires the full attention of the participant and completely determining the visual and inertial motion. While the visual-inertial nulling measure can address display factors such as the influence of FOV, resolution, etc., it is not suited to the study of interactive virtual environments. An important question, therefore, is whether less burdensome perceptual measures are possible which preserve the advantages of the visual-inertial nulling measure.
A possiblity for such a measure is suggested by Pilot Study AIIIP2 (see Appendix D). The ``induced motion'' phenomenon found in Pilot Study AIIIP2 was discussed in Section 7.4. Briefly, a grid in the visual background of the driving simulator which was stationary with respect to the laboratory appeared to rotate in the opposite direction when the CI in the foreground turned.
An interpretation of this phenomenon is that the selected rest frame was determined more by the CI than by the background grid or by the (lack of) inertial cues. The relative motion between the CI and the background grid was therefore perceived as a motion of the background grid. According to the presence hypothesis, the degree to which the selected rest frame is determined by the CI is related to the level of presence in the CI. This suggests that a perceptual presence measure may be constructed by measuring the amount of induced motion of the background grid.
A possible technique for doing so is as follows. Give participants a dial to turn, and instruct them to adjust the dial so that the background grid appears to be stationary during turns in the simulator. The dial has the effect of introducing a counter-rotation of the background grid to balance out the induced motion. The magnitude of counter-rotation needed to make the grid appear to be stationary is the perceptual presence measure.
The dial could be mounted on a navigational device for the simulator scene (e.g., a steering wheel) where it would be easy for the participant to adjust. If the level of presence is fairly constant the dial might not require updating frequently, in which case it would require little attention from the participant.