Abstract
Observers often fail to notice dramatic changes to their environment, and such failures can have real-world consequences (e.g., failing to detect hazards while driving). We examined whether 16-hours of adaptive training on a simple change detection task could improve older adults' ability to detect driving-related changes and to perform other cognitive tasks. Participants saw an initial display of 3 or 5 objects, followed by a mask (80 ms) and then a test display in which one object from the initial display was replaced by a novel object. Participants were asked to identify the changed object. We adjusted the initial display time (i.e. encoding time) individually for each subject to hold performance at 75% accuracy. A control group played online card games for 16 hours. Both groups completed two additional tasks, once before and once after training: (a) a change detection task similar to the trained task but using different stimuli (Luck & Vogel, 1997), and (b) a flicker change detection task using driving scenes (Pringle et al., 2001). Training enhanced change detection performance with the trained objects; following training, subjects required significantly shorter presentation durations (mean improvement = 26 ms at set size 3, 212 ms at set size 5) in order to achieve 75% accuracy. However, training led to no differential improvement, compared to the control group, on either of the untrained tasks. That is, training led to improvements on the trained task but did not generalize to other change detection tasks. Although the ability to detect changes might be trainable, training benefits apparently do not extend beyond the specifically trained task.