Abstract
Illusory self-motion (or “vection”) in virtual environments is often accompanied by feelings of motion sickness (or “cybersickness”), limiting the accessibility of virtual reality (VR) applications. We investigated the effects of three factors (speed, body posture, and alignment between the real and virtual reference frames) on vection, cybersickness, and sense of presence in a VR driverless car simulation. We used an HTC Vive head-mounted display to present participants with laps generated in iRacing (a driving simulator) and manipulated driving speed (60mph vs. 120mph), body posture (sitting upright or lying reclined 30 degrees backward from vertical), and the virtual reference frame (aligned with real world, or misaligned by 30 degrees in pitch). Participants (N = 24) experienced eight 60-second laps around a virtual race track (one lap per condition) and reported vection (magnitude, onset time, duration), cybersickness, and presence after each lap. 3-way repeated measures ANOVAs indicated a significant main effect of driving speed on vection magnitude, vection onset time, vection duration, and presence (Fs > 7.67, ps < .05). Vection was stronger in magnitude, faster in onset time, and longer in duration in the fast driving conditions. Participants also felt more present in the fast driving conditions. Furthermore, 3-way ANOVAs indicated a significant interaction between body posture and the virtual reference frame on vection magnitude (F = 4.85, p < .05), such that vection was stronger when the virtual reference frame was parallel with the participants' viewpoint. Additionally, 3-way ANOVAs indicated a significant interaction between driving speed and reference frame alignment on cybersickness, such that driving speed had a greater effect on cybersickness when the virtual and real reference frames were aligned (F = 6.44, p < .05). Notably, neither driving speed nor reference frame alignment had a main effect on cybersickness.