Abstract
The question that we tried to answer is how best to construct a visual message to be seen by passing motorists, albeit briefly, on the so-called ‘variable message sign’ (VMS). Optimizing the manner in which information is presented would seem to be useful to ensure that a driver is capable of reading and remembering the details within the brief time window during which the sign is readable. (An alternative strategy, forcing the driver to slow to increase that time, tends to increase congestion). Ten observers (Os) were asked to recall the information presented on a series of computer-simulated CMS signs (each of a total possible three lines of 16 characters) displaying unpredictable but comprehensible messages (e.g. 1-800-CALL-CHP; BLK HONDA SUV; LIC 4KGE268). Computer generated movies containing a simulated CMS sign were presented on a computer monitor as if it were being approached, then passed, at 96 KPH. Simultaneously, Os were engaged in a task to load attention as required in driving. They monitored 3 distinct eccentric loci on the same monitor and were asked to indicate when a target letter occurred at one. Independent CMS message variables included modifications in the justification of the text, the presentation of license plate information and notification phone number, the number of frames used, and the length of time each frame of multiple frame messages was presented. Accuracy of recall upon message extinction was scored (e.g. vehicle type, color, each character of a license plate number, etc. were assigned points). Displaying the information with left-justified text was best (p= 0.04, t-test). Characters ‘CA’ in advance of the license numbers made the license number harder to recall (p= 0.03; t-test). Messages are most effective when limited to a single frame. The second frame substantially reduced the observer's ability to reproduce information (p = 0.01; t-test). Little things seem to matter.
University of California Tranbsportaiton Center, Caltrans