Department of Psychology, University of Victoria, Victoria, BC, Canada.
Cogn Emot. 2012;26(6):961-77. doi: 10.1080/02699931.2011.630933. Epub 2012 Jan 24.
In the face literature, it is debated whether the identification of facial expressions requires holistic (i.e., whole face) or analytic (i.e., parts-based) information. In this study, happy and angry composite expressions were created in which the top and bottom face halves formed either an incongruent (e.g., angry top + happy bottom) or congruent composite expression (e.g., happy top + happy bottom). Participants reported the expression in the target top or bottom half of the face. In Experiment 1, the target half in the incongruent condition was identified less accurately and more slowly relative to the baseline isolated expression or neutral face conditions. In contrast, no differences were found between congruent and the baseline conditions. In Experiment 2, the effects of exposure duration were tested by presenting faces for 20, 60, 100 and 120 ms. Interference effects for the incongruent faces appeared at the earliest 20 ms interval and persisted for the 60, 100 and 120 ms intervals. In contrast, no differences were found between the congruent and baseline face conditions at any exposure interval. In Experiment 3, it was found that spatial alignment impaired the recognition of incongruent expressions, but had no effect on congruent expressions. These results are discussed in terms of holistic and analytic processing of facial expressions.
在面部表情识别的研究中，人们争论的焦点是识别面部表情是需要整体（即整个面部）还是分析（即基于部分）的信息。在本研究中，我们创建了快乐和愤怒的复合表情，其中面部的上半部分和下半部分形成了不一致（例如，愤怒的上半部分+快乐的下半部分）或一致的复合表情（例如，快乐的上半部分+快乐的下半部分）。参与者报告了目标面部上半部分或下半部分的表情。在实验 1 中，与基线的孤立表情或中性脸条件相比，在不一致条件下目标半脸的识别既不准确又较慢。相比之下，在一致条件下没有发现差异。在实验 2 中，通过呈现 20、60、100 和 120 毫秒的面部图像来测试暴露持续时间的影响。在最早的 20 毫秒间隔就出现了不一致面孔的干扰效应，并持续到 60、100 和 120 毫秒的间隔。相比之下，在任何暴露间隔下，一致和基线面孔条件之间都没有差异。在实验 3 中，发现空间对准会损害对不一致表情的识别，但对一致表情没有影响。这些结果从面部表情的整体和分析处理的角度进行了讨论。
