The impact of task workload on cybersickness

Abstract

This study explored the impact of task workload on virtual reality (VR) cybersickness. Cybersickness is a negative side effect of using VR to which many users are susceptible. Previous research on the impact on task workload on cybersickness has yielded no consistent relationships, but given that task workload requires attentional resources, it is worth further investigation of how a demand on attention might increase or decrease cybersickness. In this study, mental workload of participants (N = 151) was modified in three different task groups within the same virtual environment (VE). The Cybersickness Corn Maze VR testbed contained cybersickness-inducing stimuli and tasks with varying workload. The 0-Back group used a controller to select an object as a visual attention task. The 2-Back group performed the 2-Back memory detection task, using a controller to collect objects that matched the object presented two objects ago. The No-Task group passively moved through the environment and was not given a controller. Workload, cybersickness, dropout rate, presence, and task accuracy were compared across groups. Workload was found to be statistically significantly different in each group: highest in the 2-Back group, medium in the 0-Back group, and lowest in the No-Task group, validating the task design. Cybersickness in the 2-Back group was significantly higher than in the No-Task (140% higher) and 0-Back (54% higher) groups measured by the change in simulator sickness questionnaire (SSQ) total severity (Post SSQ - Pre SSQ). The rate of participants’ dropout due to cybersickness was significantly higher in 2-back (33%) as compared with 0-Back (10%), but 0-Back and No-Task dropout rates were not significantly different. These results indicate that 1) task workload affects cybersickness and 2) its effect could be based on a threshold of workload. Presence increased with the addition of a task but plateaued between the 0-Back and 2-Back groups, suggesting that presence can be affected by task workload but only to a certain extent. Task accuracy was shown to negatively correlate with cybersickness within the task groups. A relationship between workload and cybersickness was found and warrants further research into these concepts. This work highlights the need for task workload and attention to be studied as components of the mechanisms underlying cybersickness.