|
nif:isString
|
-
The experiments were approved by the Human Ethics Committee of the University of Wollongong (approval number HE10/120). All subjects participated voluntarily and gave informed written consent. Research was conducted in accordance with the principles expressed in the Declaration of Helsinki.
Participants were 10 undergraduate and graduate students (the undergraduate students received course credit for their participation), and the two authors. All had normal or corrected-to-normal vision and reported no vestibular disorders or deficits.
The stimuli were programmed on a Mac Pro computer (Mac Pro 3.1, Quad-Core Intel Xeon 2.8 GHz) using Matlab Version R2009b and Psychtoolbox [57], [58], and displayed using a Mitsubishi Electric colour data projector (Model XD400U) back-projected onto large (1.48 m wide by 1.20 m high) screen mounted on the lab wall. Subjects viewed stimuli through black-lined viewing tube fronted by a rectangular black cardboard frame, at a distance of 1.53 m from the screen, to give a field of view of 44 degrees horizontally and 26 degrees vertically. Stimuli were random clouds consisting of 1000 blue circular dots, moving in a radially expanding fashion (see Demo Movies S1 and S2), within a virtual cloud of dots simulating a “world” 30 by 30 by 80 m. Subjects were seated on a raised chair in front of the viewing tube. Their eye-height on this chair coincided with the focus of expansion of the optic flow display. During the experiment the windowless room was darkened and any external sources of light were minimised (e.g. by turning off the external monitor, etc).
Participants were asked to compare smooth radially-expanding flow (travelling at a simulated MID speed of 4 m/s) with radially-expanding flow that contained a vertically oscillating component (oscillation magnitude was 1/8 of the MID speed, or 0.5 m, and the frequency was 2 Hz). Each interval was a 1-second long motion display, and there was a 300 ms gap between stimulus presentations. The stimulus length of 1 second was specifically chosen as being too short to induce vection, as it is well established that it is not possible to induce illusory vection in stationary observers with display presentations under 3 seconds [7], [43]. Participants were specifically instructed to ignore the vertical motion and just match the stimuli for MID. Presentation was in a two-interval forced choice paradigm, with two randomly interleaved staircases where the speed of the smooth motion was manipulated using QUEST [59]. Responses were collected using a mouse button press, with participants responding with the left mouse button if the first interval looked faster, and the right button if the second looked faster. Each trial proceeded after a decision for the previous trial had been made. Participants each ran two blocks of two interleaved staircases comprising 25 trials each, giving a total of 100 trials for each participant. Results were then fitted with cumulative Gaussian psychometric functions using custom Matlab code, to give a value for each subject's point of subjective equality (PSE) for the speed of smooth radial flow that matched the perceived speed of the oscillating flow. This value was then used in Experiment 1.2 to present individually speed-matched displays for each participant.
Experiment 1.2: Vection measurements for smooth, oscillating and speed-matched stimuli: Stimuli and apparatus were as described above, but displays were now presented for longer periods of time, 30 s per trial. Participants were given a throttle control device [CH Pro USB throttle] and, after being given a basic description of vection, were asked to move the throttle forwards, if and when they felt that they were moving, to rate the extent to which they felt they were moving (and specifically not the speed of their self-motion), and to move it back if they felt they were moving less or had stopped moving; the device had tactile marking points (small raised bumps at 0, 50 and 100% positions), to assist participants in rating vection strength. The computer was programmed to require the throttle to be reset to 0 before the next trial could proceed. Latency for experiencing vection was calculated as the number of seconds before the throttle value reached a threshold of 5%; throttle maximum was defined as the maximum value that the throttle reached during each 30 second trial (see Figure 4). After each trial, participants were also asked to also give a verbal rating of their vection experience, from 0 (no self-motion) to 10 (complete self-motion); this was followed by a blank period of 5 seconds to help reduce any residual effects of adaptation. Three types of trials were randomly interleaved: smooth radial motion at 4 m/s (‘slow’), smooth radial motion moving at the individually-chosen speed that matched the perceived MID speed of the oscillating stimuli (‘speed-matched’), and oscillating radial flow moving at 4 m/s and oscillating at 2 Hz, as described above (‘oscillating’). Each stimulus type was presented 4 times, and there were 2 sessions, giving a total of 8 trials per stimulus type for each participant.
Experiment 2.1: Speed comparison for jittering stimuli: Participants were 8 undergraduate and graduate students (the undergraduate students received course credit for their participation), and the two authors. All had normal or corrected-to-normal vision and reported no vestibular disorders or deficits.
These jittering radial flow displays were exactly as above, with the single exception that, instead of smooth vertical sine-wave oscillation, the stimuli were programmed to simulate random vertical viewpoint jitter, with the virtual camera moving vertically to a new, randomly-generated location every 3 frames. Since both the magnitude and the sign of this jitter varied randomly from one jittering frame to the next, it is best represented as a range of frequencies, extending from zero to the capping frequency (10 Hz) convolved with the impulse response of the display. The amplitude of this jitter was half of that reported above for the sine wave oscillation, as pilot testing showed that this jitter amplitude produced the most realistic display motion. See Demo Movie S3 for an illustration of the jittering stimulus.
Experiment 2.2: Vection measurements for smooth, jittering and speed-matched stimuli: Experiment 2.2 was run exactly as described in Experiment 1.2, with the single exception that jittering displays (as described above) were used in place of oscillating displays. Experiment 3: Speed discrimination for all MID stimuli: Participants were 8 postgraduate students and the two authors (mean age 29.1, SD 8.65; 5 males). All had normal or corrected-to-normal vision and reported no vestibular disorders or deficits.
The stimuli were smooth-moving, oscillating or jittering radial flow displays, exactly as described above, projected onto a large screen. The experiment was run as in Experiments 1.1 and 2.1, as a speed discrimination experiment, with the stimuli being presented in 1-second intervals interleaved with a 300-ms gap, and a mouse button used to provide the 2AFC response (which interval contained faster MID?). Two randomly-interleaved QUEST staircases were used, with 25 trials each, and each participant ran two blocks of each condition, giving a total of 100 trials for each speed discrimination. We ran 4 m/s smooth, 4 m/s jittering, 4 m/s oscillating, and 6 m/s smooth conditions - the faster condition was set between the averages of perceived speed increase for oscillating and jittering conditions, with the expectation that speed discrimination threshold might rise with either actual or perceived speed, which might throw some light on the results for vection.
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}