Abstract
Visual features associated with reward involuntarily capture attention. This value-driven attentional capture effect is believed to occur even when the rewarded stimuli are irrelevant to task demands. However, almost all existing studies presented the rewarded stimuli at or near potential target locations, making them spatially relevant to the observer. To more thoroughly examine the attentional priority of reward-associated stimuli independently of spatial task relevance, in three experiments we followed reward training with a rapid serial visual presentation (RSVP) task. In the RSVP task, participants identified a letter in a specified target color among a centrally-presented letter stream, while attempting to ignore a flanker distractor in the periphery. The timing of the flanker relative to the central target was manipulated. Previous research has shown that if a flanker captures attention, it will lead to greater cost in target identification when it shortly precedes the target, compared to simultaneous presentation. In all three experiments a previously rewarded flanker did not capture attention at task-irrelevant peripheral locations, while distractors in the goal-matching color did capture attention at the same eccentricity. To confirm that the color-reward association was retained during the RSVP task, Experiment 3 included interspersed blocks of an additional singleton task, a traditional test of value-driven attentional capture that presents the rewarded stimuli at potential target locations, with blocks of the RSVP task. The additional singleton task demonstrated significant capture by the rewarded stimuli throughout the session, verifying effective reward learning, despite the absence of value-driven capture in the RSVP task. The results reveal a modulating role of spatial task relevance in the attentional prioritization of reward-associated stimuli—rewarded stimuli do not capture attention at task-irrelevant locations.