Neural correlates of reward-driven attentional capture in visual search.

Event-related potentials (ERPs) were used to investigate whether a physically non-salient and task-irrelevant stimulus feature previously associated with reward can capture attention. In the training phase, participants implicitly associated a certain color with reward. In the subsequent test phase, participants searched for a uniquely shaped singleton among non-target shapes, with color completely irrelevant to the current task demand. Response time was delayed on trials wherein the target was simultaneously presented with an associated distracter, compared with trials without one. In ERPs, a lateralized reward-associated distracter elicited an N2pc component when a concurrent target was presented on the vertical meridian. The control experiment discounted the possibility that this N2pc was caused by familiarity of the reward-associated distracter. Presenting both the target and reward-associated distracter on the opposite side elicited N2pc, with the reward-associated distracter-elicited N2pc preceding the target-elicited N2pc, albeit only in trials with slow responses. These results show that the participants shifted their attention to the target only after noticing the task-irrelevant reward-associated distracter in slow trials. Therefore, task-irrelevant reward-driven salience can capture attention. Distracter positivity was observed in fast distracter-present trials presenting both a lateral distracter and a midline target. Thus, the reward-associated distracter should be actively suppressed to efficiently select the task-relevant target. Lastly, the distracter-elicited N2pc was negatively correlated with individual differences in reward drive score. This correlation may provide insight into reward sensitivity problems.