Paul Green, Eileen Hoekstra, Marie Williams,
Cathy Wen, and Kellie George
Forty-eight drivers, while seated in a mockup of a car, watched a videotape of a driver's eye view of a 25-minute trip. Simultaneously, they received route guidance and traffic information in one of the four formats: visual, visual with landmarks, auditory, or auditory with landmarks. Drivers pressed buttons when they could see the intersections in the videotaped scene referred to by the route guidance system. They also stepped on the brake pedal when a car immediately in front of them braked. In addition, subjects rated the effect that the reported traffic problems would have on their travel.
Following the test session, subjects completed questionnaires concerning the ease of use and usefulness of the route guidance system. They also indicated their preferences for all four systems and rated the difficulty of performing driving tasks. Differences in task difficulty due to interface design were small and not correlated with the results from a subsequent on-the-road experiment. Glance data in the laboratory were not correlated with glance data collected on-road for the same roads. These results, along with problems in recovering the data, suggest that the videotape method may not be a good approach for evaluating the usability of route guidance driver interfaces.
|UMTRI-93-31 Full Report (.pdf)|
Close This Window