Efficient Training Of Visual Search Via Attentional Highlighting

Human activities often involve interaction with dynamic or complex visual environments, e.g., driving an automobile or inspecting radiographic images. Knowing where to look and what features to look for in a given context is critical to expertise, but this knowledge is typically acquired by effortful and inefficient trial-and-error learning. We ask whether training of endogenous attentional control can be improved by highlighting regions of interest in a display. Participants searched displays whose global configurations predicted the location of a target. Training reduced response latencies—indicating learning of contingencies—whether during training participants effortfully searched for targets or were directed to target locations via highlighting (brief flashes). However, highlighting was effective only when the training task required no response beyond a saccade to the target. By halving training time and demanding less effort, highlighting shows practical potential and also provides insight into the interplay between exogenous and endogenous attentional control.