Gretchen Paelke, Paul Green, and Cathy Wen
This report describes a series of three efforts that were carried out to develop safe and easy to use in-car traffic information systems and guidelines, and methods for their evaluation.
Desired information elements were identified based on the literature. Several candidate designs were constructed, some of which were not developed further because of screen space constraints. Using the Keystroke and Tullis Models, designs were eliminated from further consideration or modified, and subjected to quick usability tests concerning information arrangement and retrieval.
Next, at a driver licensing office, 20 patrons were shown color copies of five alternative coding schemes. For presenting traffic information graphically, a green-yellow-red color key was best understood, and it was preferred to include actual travel speeds.
In a subsequent driving simulator laboratory experiment, 16 drivers retrieved traffic information on request. Among text-based systems, retrieval times using a graphic of highway buttons and a scrolling menu were less than those for a phone-style keypad method. Glance behavior to these displays involved several short glances. Response times for graphic systems were faster overall due to the reduced retrieval effort. Drivers typically made fewer, but considerably longer glances to the graphic system displays.
Overall, the graphic of highway buttons was significantly preferred by participants, while the scrolling menu was least preferred. The standard deviation of lane position (in a driving simulator) increased by about 3-4 inches while using the traffic information systems.
Close This Window