Abstract: Attention is automatically drawn to stimulus features previously associated with reward, a phenomenon referred to as value-driven attentional capture. Most previous studies associated a target-defining single feature(color, shape, orientation) or spatial location with reward in the training phase. However, task-irrelevant spatial context had never offered uniquely predictive information about the receipt of reward in previous studies. It is unclear whether feature and location-based value-driven attentional capture is extended to spatial context. Anderson(2014) investigated the roles of target color and spatial context in value-driven attentional capture, but in his research, neither the target color nor the context could predict reward independently. When target color and spatial context can predict reward independently, the pattern they affect attention is need to be examined. The present study used training-testing paradigm to explore the role of spatial context in value-driven attentional capture. Experiment 1 was designed to test whether context-value association can capture attention via reward-based learning. In training phase, correct response was followed by high reward when the red color target appeared in context A, and low reward feedback when it appeared in context B. In test phase, all participants completed a search task in which color was irrelevant. In some trials, the red stimulus was presented as distractor in different contexts. Experiment 2 was designed to test the interaction effect of color and context in value-driven attentional capture. In training phase, correct response was followed by high reward feedback when the X-color target appeared in context A, and low reward feedback when Y-color target appeared in context B. In test phase, X-color or Y-color stimulus was presented as distractor in different contexts. In experiment 1, we found when red stimulus was presented in the context associated with high reward, it could capture attention in test phase. The result in experiment 2 showed when integration information of color and context was associated with reward, but each information could predict reward independently, color distractor associated with high reward appeared in test phase, it could capture attention regardless of the context. In addition, we observed faster search times on high reward compared to low reward trials in training-phase of all experiments. The present findings suggested that (1) The spatial context associated with high reward can capture attention. This result provide strong evidence for classical (Pavlovian) conditioning, stimulus-reward contingencies is sufficient to obtain value-driven attentional capture; (2) When target color and context can predict reward independently, the individual uses only one reward-signaling stimuli(color), and another reward-signaling stimuli(spatial context) can be ignored. This result is consistent with the Attentional Theory of Context Processing.

Key words: value-driven attentional capture, context, color feature

摘要： 本研究采用训练-测试范式考察场景信息在价值驱动注意捕获中的作用。在实验一中，训练阶段将唯一性场景信息与奖励建立联结。测试阶段发现，只有当分心刺激出现在高奖励场景中才能捕获注意；在实验二中，训练阶段将场景和颜色信息联合但分别独立预测奖励。结果发现，高奖励颜色出现在任何场景中都能捕获注意。研究结果表明：(1)场景-价值联结能够引发注意捕获效应；(2)与场景相比，颜色特征在价值驱动的注意捕获中具有优势效应。

关键词: 价值驱动注意捕获, 场景, 颜色特征