Department of Psychology.
J Exp Psychol Learn Mem Cogn. 2020 Apr;46(4):669-683. doi: 10.1037/xlm0000749. Epub 2019 Jul 25.
Mounting evidence suggests that monetary reward induces an incidentally learned selection bias toward highly rewarded features. It remains controversial, however, whether learning of reward regularities has similar effects on spatial attention. Here we ask whether spatial biases toward highly rewarded locations are learned implicitly, or are instead associated with explicit knowledge of reward structure. Participants completed a hybrid search and choice task involving multiple targets among multiple distractors. Targets garnered varying magnitudes of reward, and participants were instructed to search for targets and guess and click on the 1 that they thought would yield the highest reward. Unbeknownst to participants, 1 side of the display offered higher reward than the other. We measured the spatial bias for targets on the high-reward side of the screen and probed explicit awareness via a multiquestion interview. Participants who were aware of the reward structure (N = 48) showed a selection bias for targets appearing on the high-reward side of the screen. Contrary to previous findings, unaware participants (N = 24) showed only a significant central bias, despite spending just as much time on the task. The strong association between explicit awareness and reward-driven spatial attention in this paradigm suggests that instead of directly affecting the attentional priority map, probabilistic spatial reward learning more frequently affects attention indirectly by modulating task goals. (PsycInfo Database Record (c) 2020 APA, all rights reserved).
越来越多的证据表明,货币奖励会引起偶然习得的选择偏差,偏向于高奖励特征。然而,奖励规律的学习是否对空间注意力有类似的影响仍存在争议。在这里,我们想知道对高奖励位置的空间偏向是通过内隐学习还是与对奖励结构的明确知识有关。参与者完成了一个混合搜索和选择任务,涉及多个干扰项中的多个目标。目标获得不同大小的奖励,参与者被指示搜索目标并猜测并点击他们认为会产生最高奖励的目标。参与者不知道的是,显示的一侧比另一侧提供更高的奖励。我们测量了屏幕高奖励侧目标的空间偏向,并通过多项问题访谈来探测明确的意识。意识到奖励结构的参与者(N=48)表现出对屏幕高奖励侧目标的选择偏差。与之前的发现相反,尽管参与者(N=24)在任务上花费的时间相同,但无意识参与者仅表现出显著的中央偏向。在这个范式中,明确意识与奖励驱动的空间注意力之间的强烈关联表明,概率空间奖励学习不是直接影响注意力优先级图,而是更频繁地通过调节任务目标间接影响注意力。