The Unexpected Vection Hypothesis: Exploring New Relationships Between Vection, Postural Instability, and Cybersickness in Virtual Reality

<p dir="ltr">Head mounted display Virtual Reality (HMD-VR) systems use advanced computer technology to create immersive, simulated environments that allow users to experience the sensation of being physically present in a digital space. While VR is increasingly employed in entertainment, training, and therapeutic applications, its widespread adoption is hindered by cybersickness—a condition characterised by a range of adverse symptoms, including nausea, dizziness, and visual discomfort. Cybersickness, which shares similarities with motion sickness, is unique to virtual environments and presents a significant challenge to maintaining user comfort and immersion. One factor often argued to cause cybersickness is vection, the illusion of self-motion that occurs when large-field visual motion induces a sense of movement in a stationary observer. While vection can be a positive aspect of VR and enhancing the sense of presence and realism of virtual environments, the widespread belief that it plays a causal role in cybersickness has made it seen to be a problem that should be avoided. However, the precise nature of the relationship between vection and cybersickness remains unclear, with conflicting evidence suggesting that vection can either contribute to, have no effect on, or even mitigate cybersickness. This thesis aimed to address these inconsistences in the literature and identify exactly which types of vection (if any) contribute to cybersickness. It first outlined the ‘unexpected vection hypothesis', which proposes that cybersickness only occurs when these illusions of self-motion are perceived in ways that violate user expectations. This new theory of cybersickness was tested alongside the postural instability theory, which proposes that cybersickness occurs when users are unable to maintain stable control of their posture in response to conflicting visual/vestibular information. The thesis consisted of 4 experiments, where participants engaged with commercially available HMD-VR video games for up to 14 minutes. Across all of these experiments, the results consistently supported the notion that unexpected vection serves as a reliable and robust predictor of cybersickness occurrence and severity. In the first three thesis experiments (Chapters 2-4), active participants were instructed to play one of two video games (either Mission: ISS or Aircar), and their experiences of vection type and strength, as well as their cybersickness, were assessed. In all 3 experiments, cybersickness was found to increase with the strength of any unexpected (but not expected) vection. In the final thesis experiment (Chapter 5), participants were passively exposed to stereoscopic first-person 360° videos of pre-recorded gameplay (from Project Cars), with their experiences of vection (expected/unexpected) and cybersickness being assessed every minute throughout each exposure. Its different experimental conditions were designed to generate either more or less instances of unexpected vection. Consistent with the unexpected vection hypothesis, conditions which generated more instances of unexpected vection were also found to provoke more severe cybersickness (compared to conditions designed to favour the induction of smooth, expected vection). Importantly, reports of unexpected vection were also found to be linked to an increased likelihood of a subsequent rise in sickness severity, suggesting that unexpected vection preceded and likely caused this sickness. While postural instability also predicted sickness in some contexts, unexpected vection consistently emerged as the strongest predictor. These results challenge the assumption that vection is inherently problematic, demonstrating instead that its unpredictability—rather than its presence or strength alone—determines its impact on sickness. Managing unexpected vection—whether through anticipatory cues, improved motion predictability, or adaptive design strategies—may serve as a crucial intervention for reducing cybersickness and improving user comfort. Future work should investigate how cognitive and sensory processes interact to produce unexpected vection, explore its relationship with other motion cues, and test targeted mitigation strategies.</p>