Abstract
We have reported that the ground stimulus is more effective than the ceiling in inducing vection (Sato et al., 07 ECVP). The ground dominance effect (GDE), however, had been demonstrated only when the observer was standing upright, and it was not clear which coordinate system (e.g. retinal/world) determines the ground and the ceiling. To examine this point, in this study, we had observers evaluate vection strength induced by either the ground or the ceiling stimulus while keeping one of the following three positions; upright, upside-down, or 90 deg tilt (sideways). The ground and ceiling stimulus (random-dot pattern) were projected upper or lower 1/3 of a large screen (75 deg horizontally X 65 deg vertically). The motion of the dots simulated a forward motion of the observer (1.7 m/s). In each trial, both ground and ceiling stimulus was presented first as a standard, then, a test (either the ground or ceiling alone) was presented for 60 sec. Observers were asked to make magnitude estimation for the vection induced by the test stimulus relative to that by the standard. Observers were asked to press a button while they perceived vection to measure the latency and duration of vection. It was found that the head orientation had little effect on the GDE in vection. In an additional experiment, observers in the sideways position observed the stimulus that was also rotated sideways. That is, the orientation of the stimulus was upright according to the retinal coordinate. In this condition, although vection of normal magnitude occurred, no GDE was identified. These results, thus, indicate that the ground dominance of vection occurs according to the ground defined by the external or the world-centered coordinate instead of the retinal or the body centered coordinate.