Prediction of simulator sickness in a virtual environment

Sickness induced by Virtual Reality (VR) devices poses a genuine threat to the viability of this new technology and its potential products. If the occurrence or severity of sickness could be successfully predicted based on characteristics of an individual , atrisk users could be identified, properly warned, and, perhaps, trained in some way to reduce their risk. A Personal Computer-based VR system was used to address the prediction of simulator sickness. Phase I investigated four characteristics of an individual age, gender, mental rotation ability, and pre-exposure postural stability which were hypothesized to be predictive of sickness. Sickness measured as a function of the Total Severity score from the Simulator Sickness_, Questionnaire (SSQ) was successfully modeled on these characteristics using linear regression techniques, leading to three major findings. First, sickness as measured by the SSQ did, in fact, occur in association with exposure to VR. For 35% of the participants, this sickness involved lingering effects and/or possible delayed after-effects. Second, sickness was successfully modeled on characteristics of the individual. The developed model indicated a complicated relationship between predicted sickness and gender, age, mental rotation ability , and pre-exposure postural stability. Third, based on the model developed, sickness is not predicted to differ for gender directly but, rather, gender interacts with mental rotation ability in its effect on sickness. Phase II investigated the occurrence of ataxic decrements in postural stability. No such decrements were found to be associated with the 20-minute exposure. Thus , ataxic decrements do not appear to be associated with short exposures to low-end VR. This finding, however, may be limited to VR tasks of the type used in this study. Practical implications and areas for future research are discussed.