Pavlovian reward learning underlies value driven attentional capture

摘要

Typically, if attending a target stimulus is consistently paired with obtaining high reward, then that stimulus becomes more likely to capture attention compared with an equally salient stimulus associated with low reward. Recent evidence extends this finding by showing that task-irrelevant distractors that signal high compared with low reward availability also elicit stronger capture, even when this is detrimental for task-performance and monetary payout. This suggests that the simple correlation between stimuli and reward delivery, rather than their instrumental relationship with obtaining reward, underlies value driven attentional capture. However, in previous studies, reward delivery was never response independent, as only fast and correct responses were rewarded, nor completely task-irrelevant, as the only way to discover how much reward could be earned on a trial was to attend the reward signaling distractor. In the present study we specifically addressed whether the mere correlation between stimuli and obtaining reward, completely independent of the current task and/or response, was able to elicit value driven attentional capture. During a classical (Pavlovian) conditioning phase, twenty-four participants performed a fixation task while stimuli following high and low reward feedback were presented in the periphery. In a subsequent testing phase in which no rewards were distributed, participants performed the additional singleton task while one of the distractors was sometimes presented in the previously high or low reward-value associated color. The results revealed that high compared with low reward-value associated distractors significantly slowed response times. Furthermore we showed that value driven attentional capture was strongest directly after the reward conditioning phase and decreased over time. These results provide clear evidence that value driven attentional capture occurs following a classical conditioning procedure. This confirms and strengthens the idea that Pavlovian rather than instrumental learning of stimulus-reward contingencies underlies value driven attentional capture.