In typical dual-task driving studies, participants concurrently perform pairs of driving-related and -unrelated tasks (e.g. vehicle braking and mental arithmetic). Requiring responses to both may implicitly equate their importance. In real-life driving, however, the potential for collision dictates that a concurrent task should be assigned far lower priority than driving. To better reflect naturalistic driving conditions, we not only instructed participants to assign maximum priority to braking in a simulated driving task, but also encouraged them to ignore the concurrent task altogether on dual-task trials. Despite these instructions, responses to the concurrent task often preceded braking, which suffered from dual-task interference.We also found that redundant signals to the lead vehicle’s brake lights resulted in faster braking responses and an increased likelihood that the braking response would occur first. The results are consistent with the Central Bottleneck (CB) model of dual-task interference and may help guide the design of driver-assistance systems. (Author/publisher)
Abstract