Abstract
The incidence of automotive crashes is particularly high among drivers under the age of 25 (Evans, 2004; Williams & Carsten, 1989) and over the age of 65 (Tefft, 2008; Evans, 2004). An important issue for driving safety is identifying potential ways to reduce crash risk by examining drivers’ ability to detect impending collisions. Perceptual learning (PL)---improvements on a perceptual task as a result of repeated exposure---has been shown to improve collision detection in both older and younger adults (Deloss, Bian, Watanabe & Andersen, 2015; Lemon, Deloss & Andersen, 2017). An important component of PL training is attention, which has been found to improve perceptual performance for a variety of perceptual tasks (see Byers & Serences, 2012). The present study investigated the effect of attention and PL on collision detection for both older and younger drivers in a dual-task driving paradigm.
Drivers were presented with a computer-simulated roadway scene and maintained within-lane vehicle steering while also identifying which object among a number of objects (number of objects; 2,4,8) will collide with the driver. During training, drivers were trained over several days with a number of objects (2,4,8). Drivers were either presented with an endogenous cue (identifying the visual field location of the collision object) or a neutral cue. The results indicated overall decreased detection performance (lower accuracy and greater RT) with an increase in the number of objects. PL resulted in improved collision detection performance for both older and younger drivers. In addition, greater accuracy and RT were found when an endogenous attention cue was present with improved performance for both younger and older drivers. These results indicate the benefits of training and attention in collision detection.