A Pool of Pairs of Related Objects (POPORO) for Investigating Visual Semantic Integration: Behavioral and Electrophysiological Validation
A Pool of Pairs of Related Objects (POPORO) for Investigating Visual Semantic Integration: Behavioral and Electrophysiological Validation
复制标题
DOI：
10.1007/s10548-011-0216-8
复制
发表时间：
2012-07-01
期刊：
影响因子：
2.7
通讯作者：
Busch, Niko A.
中科院分区：
文献类型：
--
作者：
Kovalenko, Lyudmyla Y.;Chaumon, Maximilien;Busch, Niko A.
言语和视觉刺激的语义加工已经在语义违反或语义启动范式中进行了研究，其中刺激与先前建立的语义上下文相关或不相关。语义启动的一个标志是N400事件相关电位（ERP）--一种对语义无关的目标刺激更负的ERP偏转。大多数调查N400和语义整合的研究都使用了口头材料（单词或句子），标准化的刺激集与规范的语义相关性已出版的口头，但不是视觉材料。然而，视觉对象（相对于单词）的语义加工是视觉认知研究中的一个重要问题。在这项研究中，我们提出了一套800对语义相关和不相关的视觉对象。这些图像的语义相关性由132名参与者的样本进行了评级。此外，我们分析了低级别的图像属性，并根据这些功能匹配的两个语义类别。一项ERP研究证实了这一图像集的适用性，唤起了强大的N400效应的语义整合。此外，使用单次试验数据的一般线性建模方法，我们还证明了低级别的视觉图像属性和语义相关性实际上只有最低限度的重叠。图像集可从作者的网站下载。我们期望图像集将有助于研究视觉刺激的语义和上下文处理机制。
Semantic processing of verbal and visual stimuli has been investigated in semantic violation or semantic priming paradigms in which a stimulus is either related or unrelated to a previously established semantic context. A hallmark of semantic priming is the N400 event-related potential (ERP)-a deflection of the ERP that is more negative for semantically unrelated target stimuli. The majority of studies investigating the N400 and semantic integration have used verbal material (words or sentences), and standardized stimulus sets with norms for semantic relatedness have been published for verbal but not for visual material. However, semantic processing of visual objects (as opposed to words) is an important issue in research on visual cognition. In this study, we present a set of 800 pairs of semantically related and unrelated visual objects. The images were rated for semantic relatedness by a sample of 132 participants. Furthermore, we analyzed low-level image properties and matched the two semantic categories according to these features. An ERP study confirmed the suitability of this image set for evoking a robust N400 effect of semantic integration. Additionally, using a general linear modeling approach of single-trial data, we also demonstrate that low-level visual image properties and semantic relatedness are in fact only minimally overlapping. The image set is available for download from the authors' website. We expect that the image set will facilitate studies investigating mechanisms of semantic and contextual processing of visual stimuli.