L. Morrison, B. Weaver, N.W. Mullen, M. Bédard
We examined the feasibility of using electrooculography (EOG) to monitor eye movements during simulated driving. We created three versions of a driving scenario that differed only in terms of how navigation instructions were provided. Two versions included visual navigation instructions, such as one would get from a global positioning device. In one version, the visual instructions appeared in the lower right corner of the middle screen (GPS group); and in the other, the instructions appeared in the centre of the middle screen, such that drivers did not have to move their eyes from the road to view the instructions. The third version included auditory navigation instructions only. We hypothesized that when visual navigation instructions appeared in the lower right corner of the screen, (1) more time would be spent glancing away from the road, and (2) that this would translate into poorer driving performance. We measured glance presence, and calculated glance latency and length during the three seconds following the onset of 12 randomly selected visual instructions. During this time, participants in the GPS group looked away from the road to the right significantly more often than those in the other two groups. As a result, these participants spent significantly more total time looking away from the road during the drive when compared to the other two groups. Groups did not differ significantly on any of the individual driving mistake categories, or on the total number of driving mistakes. Electrooculography does appear to be a feasible and affordable way to measure eye movement during simulated driving. Our results confirmed that participants spent more time glancing away from the road to the right when visual instructions were given in the lower right hand corner of the screen. Though electrooculography does not provide the same amount or quality of data as head-mounted eye trackers and multiple camera systems, it does yield sufficient data to address questions such as the ones posed in this study.
Keywords: glance analysis electrooculography; simulated driving; distraction; GPS