|
Abstract
|
Auditory warnings are a particularly useful means to signal an imminent collision. There are few other ways to inform a driver quickly -- and without involving a redirection of gaze -- that an immediate response is urgently needed.
The acquisition, response speed, and error rates of three suites of collision warning sounds were investigated to evaluate the effect of sound alteration on responding. In each suite, four sounds were pictorially associated with four collision scenarios. Suite A included two natural sounds, and two artificial sounds semantically associated with one of four crash scenarios; Suite B was a variant of A, altered to reduce perceived urgency; Suite C was a set of abstract sounds constructed to vary in urgency and matched to the subjective urgency of each scenario. For each suite, subjects first learned to associate the suite’s warning sounds with an assigned crash scenario to an established criterion. This was followed by reaction time trials in which a sound was played and subjects quickly identified the scenario associated with the sound. For both young and old subjects, Suite A produced the shortest reaction times and fewest trials to criterion, suggestive of the response efficiencies reported for auditory icons. In contrast, the sounds used in Suite B, while variants of Suite A, were most difficult to learn and were not different from Suite C with respect to error rates and reaction time. It is suggested that even relatively minor alterations of a warning sound can result in marked differences in acquisition and performance. |