Hilchey Matthew D, Leber Andrew B, Pratt Jay
Department of Psychology, University of Toronto, 100 St. George St, Toronto, Ontario, M5P 3G3, Canada.
Department of Psychology, The Ohio State University, 225 Psychology Building, 1835 Neil Avenue, Columbus, OH, 43210, USA.
Atten Percept Psychophys. 2018 Aug;80(6):1362-1374. doi: 10.3758/s13414-018-1550-7.
In common "attention" tasks, which require stimulus-identity processing prior to the formation of a speeded key-press response, spatial priming effects depend on response repetition. Typically, the repetition of a stimulus location is advantageous when the prior response repeats, but disadvantageous or inconsequential when the prior response changes. This link between responding and space makes it difficult to draw inferences about attentional bias from two-choice key-press tasks. Instead, the findings are accounted for by episodic retrieval theories, which argue that the response associated with a prior stimulus location is retrieved when a later stimulus occupies its space. This retrieval operation is advantageous if the prior response is needed but not otherwise, which explains typical patterns. This perspective motivated us to evaluate whether spatial priming effects in the visual-search literature depend critically on response repetition. To assess this, we reevaluated a series of experiments recently published by Tower-Richardi, Leber, and Golomb (Attention, Perception, & Psychophysics, 78(1), 114-132, 2016). Their goal was to determine the reference frame of spatial priming across visual search displays. Reassessment reveals that spatial priming was strongly dependent on response repetition when spatiotopic, retinotopic, and object-centered reference frames were perfectly confounded. However, when eye movements were made to dissociate the spatiotopic and object-centered reference frame from the retinotopic reference frame, spatial priming was positive and unaffected by response repetition. The findings demonstrate that at least two distinct processes factor into spatial priming across visual searches, which occur at different levels of representation.
在常见的“注意力”任务中，这类任务需要在快速按键反应形成之前进行刺激识别处理，空间启动效应取决于反应重复。通常，当先前的反应重复时，刺激位置的重复是有利的，但当先前的反应改变时则是不利的或无关紧要的。反应与空间之间的这种联系使得从二选一按键任务中推断注意力偏差变得困难。相反，这些发现可以用情景检索理论来解释，该理论认为，当后来刺激占据先前刺激的空间时，与先前刺激位置相关的反应会被检索出来。如果需要先前的反应，这种检索操作是有利的，否则则不然，这就解释了典型的模式。这一观点促使我们评估视觉搜索文献中的空间启动效应是否严重依赖于反应重复。为了评估这一点，我们重新评估了Tower-Richardi、Leber和Golomb最近发表的一系列实验（《注意力、感知与心理物理学》，78(1)，114 - 132，2016）。他们的目标是确定视觉搜索显示中空间启动的参照系。重新评估发现，当空间位置、视网膜位置和以物体为中心的参照系完全混淆时，空间启动强烈依赖于反应重复。然而，当通过眼动将空间位置和以物体为中心的参照系与视网膜参照系区分开时，空间启动是正向的且不受反应重复的影响。这些发现表明，在视觉搜索中的空间启动至少有两个不同的过程因素，它们发生在不同的表征水平上。
