Abstract
Various studies have shown that reward has a strong impact on attention deployment in visual search tasks; features associated with high reward are more likely attended in subsequent trials than features associated with low reward. The present study focussed on the role that reward plays for the processing of irrelevant distractors. We used a visual search task in which additional singletons (salient distractors) were associated with high or low rewards throughout the task and investigated which neural mechanisms (target enhancement / distractor suppression) are responsible for prioritized processing of distrators associated with reward. To that end, we used sub-components of the N2pc in the event-related EEG: the NT (target negativity) and ND/PD (distractor negativity/positivity). We also varied the context homogeneity in order to render visual search more (heterogeneous) or less (homogeneous) prone to disruption from a non-target color singleton. Participants were given a monetary reward contingent on the color of the distractor after each trial. High-reward distractors delayed responses in heterogenous contexts, but not in homogenous contexts.For both heterogeneous and homogeneous contexts, the NT, indicative of attention deployment to the target, was smaller when the target was presented together with a high-reward distractor, indicating that high-reward distractors interfere with target enhancement, regardless of context. In heterogeneous contexts, high-reward distractors elicited a larger ND component, indicating greater attentional capture, and a larger PD component, indicating more effort to suppress distractors. In homogenous contexts, ND and PD were not modulated by reward, suggesting no reward-related differences in attentional capture or suppression. In sum, the results showed that distractor suppression is only affected by reward if the context makes visual search vulnerable to attentional capture. Target enhancement, however, is always disrupted by a high-reward distractor, regardless the context.
Meeting abstract presented at VSS 2016