Abstract
A rapidly expanding literature has shown that monetary reward modulates visual attention, often in the absence of explicit instructions. A tenet is that the effect is implicit, with subjects unable to report the relative values of rewarded stimuli. Yet conditions under which reward implicitly guides spatial attention are unclear. Here we tested the effects of monetary reward on the allocation of spatial attention in visual search under incidental learning conditions. We compared these effects to intentional learning of reward and to location probability, an established form of implicitly learned attention. Participants searched for the letter T among letter Ls to uncover monetary reward hidden behind the T. Although the T was randomly located, monetary reward was three times higher in one visual quadrant than in any of the other visual quadrants. Contrary to recent findings, participants did not develop an implicit spatial bias toward the high-reward quadrant. Search RT was equivalent for targets in the high-reward and low-reward quadrants. These results were found both when reward was given without considering RT and when reward was given only when RT was faster than a stringent deadline. However, monetary reward became a powerful source of top-down attention under explicit learning conditions. Specifically, when participants were told that one quadrant was associated with greater reward, those who successfully identified the high-reward quadrant showed faster RT for targets in the high-reward quadrant. Thus, locations associated with higher monetary reward were prioritized under explicit but not implicit learning conditions. In contrast, locations that more frequently contain the search target received priority even in the absence of monetary reward or explicit awareness. We conclude that experience-driven attention is a dual-system supported by explicit goals and implicit learning. Monetary reward affects spatial attention primarily by changing explicit goals and occurs under more limited conditions than previously thought.
Meeting abstract presented at VSS 2015