The best way to assess visually induced motion sickness in a fixed-base driving simulator

Abstract

ObjectiveDriving simulator usage is becoming more widespread, yet many users still experience substantial motion sickness-like symptoms induced by optical flow, called visually induced motion sickness (VIMS). The Fast Motion sickness Scale (FMS) allows for continuous on-line assessment of VIMS. Using mixed models for ordinal data, this study investigated how to optimally analyze FMS data, and then used the resulting models to examine the development of symptoms over time in detail. Additionally, the study explored the impact of specific VIMS-inducing road elements. MethodsTwenty-eight healthy young adults without prior simulator experience completed six courses on two days in a fixed-base driving simulator. VIMS severity was reported every minute using the FMS. Each course included two road elements designed to induce VIMS. The data was analyzed using cumulative link mixed models. ResultsThe FMS data deviated clearly from a normal distribution. Treating FMS data as ordinal led to preferable models compared to models assuming interval scale. VIMS increased within each drive and over consecutive courses, but decreased between two days separated by a week (adaptation). Adaptation was attributable to less pronounced symptom increases on the second day, both within each course and between consecutive drives. VIMS increases within each drive were less pronounced during later courses of each day (habituation). Participants differed both in general symptom levels and in their progressions of VIMS over time. Additionally, VIMS-inducing road segments could be identified as leading to higher probabilities of symptom increases. ConclusionMixed models analyses of FMS data from repeated VIMS measurements can benefit from taking deviations from normal distribution and interval scale into account. The gained insights into habituation and adaptation processes, as well as into the impact of specific road elements, can help in planning and conducting future driving simulator experiments.