Suppr超能文献

选择历史和任务可预测性决定了注意力控制中的精度预期。

Selection history and task predictability determine the precision expectations in attentional control.

机构信息

Cognitive Neuroscience of Perception and Action, Philipps-University of Marburg, Marburg, Germany.

出版信息

Psychophysiology. 2023 Jan;60(1):e14151. doi: 10.1111/psyp.14151. Epub 2022 Aug 10.

Abstract

Predictive processing frameworks have demonstrated the central role that prediction plays in a range of cognitive processes including bottom-up and top-down mechanisms of attention control. However, relatively little is understood about how predictive processes interact with the third main determinant of attentional priority - selection history. In this experiment, participants developed a history of either color or shape selection while we observed the impact of these histories in an additional singleton search task using behavioral measures and ERP measures of attentional control. Throughout the experiment, participants were encouraged to predict the upcoming display, but prediction errors were either high or low depending on session. Persistent group differences in our results showed that selection history contributes to the precision weighting of a stimulus, and that this is mediated by overall prediction error. Color-singleton distractors captured attention and required greater suppression when participants had a history of color selection; however, these participants gained large benefits when the upcoming stimuli were highly predictable. We suggest that selection history modulates the precision expectations for a feature in a persistent and implicit way, producing an attentional bias that predictability can help to counteract, but cannot prevent or eliminate entirely.

摘要

预测加工框架表明,预测在包括自下而上和自上而下的注意控制机制在内的一系列认知过程中起着核心作用。然而,对于预测过程如何与注意优先级的第三个主要决定因素——选择历史相互作用,人们的了解相对较少。在这项实验中,参与者在进行颜色或形状选择时形成了一种历史,而我们则使用行为测量和注意控制的 ERP 测量来观察这些历史在额外的单一搜索任务中的影响。在整个实验过程中,我们鼓励参与者预测即将出现的显示,但根据会话,预测错误要么高要么低。我们的结果中持续存在的组间差异表明,选择历史有助于刺激的精度加权,并且这是由整体预测误差介导的。当参与者有颜色选择历史时,颜色单一的分心物会吸引注意力并需要更大的抑制;然而,当即将出现的刺激高度可预测时,这些参与者会获得很大的好处。我们认为,选择历史以持久和隐含的方式调节了对特征的精度期望,产生了一种注意偏差,可预测性可以帮助抵消,但不能完全预防或消除。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验