The study reported in this paper investigated the effect of a simultaneous secondary task on subject's ability to judge time-to-collision (TTC). Using the Leeds Driving simulator, subjects watched as they appeared to approach, at one of two constant speeds, a texture-less distant object along a flat straight road in a randomly undulating scene. At predetermined points the subject was instructed to look at an in-car display, at which point the approach to the target object was no longer presented. The subject's primary task was to decide when the vehicle in which he or she was travelling would collide with the target. Three secondary task conditions, using the in-car display were studied: (1) where the subject had to perform a difficult signal detection task; (2) an easy signal detection task; and (3) no secondary task during the TTC estimation phase. The results showed the expected difference in ability to perform the easy and difficult secondary task, but an increased collision rate (i.e. overestimation of TTC) and reduced TTC error in these secondary task conditions. The fact that these tendencies in TTC estimations also reflected the difficulty level of the secondary task suggests that what some have regarded as `direct' perception is disrupted by cognitive load. The results obtained are interpreted as showing that the TTC task depend in part on prospective duration estimation, which is itself sensitive to ongoing cognitive load. The use of the TTC estimation task as a means of assessing the impact of distraction from in-car devices is discussed. (A)
Abstract