认知与发展心理学研究室知识库

    人们可以相当精确地提取集合内多个体的平均表征，但在整体编码过程中，平均表征与个体表征到底是什么关系存在着争议。有研究发现人们可以提取到精确地平均表征，但个体表征的精确性较差;也有研究发现还可以形成比较精确的个体表征。这种争议涉及平均表征加工机制的理论争论一一基于分布式注意的整体平行加工，还是基于集中注意的个体抽样加工。基于视知觉学习的层级和反转层级理论，本文从层级加工的视角研究平均表征和个体表征之间的关系，初步整合平均表征的两种理论，有助于理解平均表征的加工机制。
    研究一从情感加工的视角，采用简单暴露范式变式，结合行为实验、ERP技术和眼动技术等多种手段，考察平均表征与个体表征的关系。结果发现，在整体编码的早期，情绪平均表征与个体表征的加工可以共存。
    研究二采用预览搜索范式，进一步考察注意抑制对平均表征和个体表征加工的影响。结果发现，在注意抑制的情况下，情绪平均表征与个体表征也可以共存。通过调控注意抑制程度，我们发现，情绪平均表征和个体表征在早期均可被加工但加工程度不同，但相比于平均表征，个体表征更不稳定。
    研究三从注意资源分配出发，考察集合不同呈现时间、集合内不同容量对平均表征与个体表征关系的调控。结果发现，短时呈现条件下，情绪平均表征的精确性较高，个体表征精确性较低，二者竞争中又有一定的共存关系;长时呈现条件下，情绪平均表征和个体表征精确性均较高，二者共存。
    我们发现，平均表征与个体表征的竞争与共存呈动态变化。在意识注意之前，个体表征同平均表征的加工可以共存，但在意识注意阶段，受注意资源限制，平均表征与个体表征表现出竞争关系。随着平均表征与个体表征获得的注意资源的增加，两者又表现出共存关系。我们认为，平均表征与个体表征的加工可能是并行层级加工的过程。
    本研究采用多种实验范式和技术手段考察了平均表征与个体表征的关系，丰富和发展了平均表征的理论。

    People can accurately extract average representation from multiple objects.However, there is a hot debate on the relationship between average representation and individual representations during ensemble coding. It has been found that individuals can accurately extract average representation at the expense of the accuracy of individual representations. But recently, there is growing evidence that observers can extract both average representation and individual representations accurately. The debate involves two theoretical arguments about the processing mechanism of average representation: the global parallel processing based on distributed attention, and the individual sampling processing based on focused attention. From a hierarchical perspective, based on the hierarchy and reverse hierarchy theory of visual perceptual learning, this study investigate the relationship between average representation and individual representations, and preliminary integrate the two theories of average representation, which contribute to understand the processing mechanism of average representation.
    In study 1，we used mere exposure paradigm variants to examine the relationship between average representation and individual representations with multiple indicators, such as behavioral performance, temporal progression (ERP techniques), and eye movement patterns (eye-tracking technique). Results showed that, average emotion representation coexisted with individual representations during the early stage of ensemble coding.
    In study 2, we adopted preview search paradigm to examine the effects of attention suppression on average representation and individual representations.Results showed that, both average emotion representation and individual representations could be processed, but to a different degree, and compared to average emotion representation, individual representations was more unstable.