Abstract
Research has shown that people perceive facial expressions of a single face holistically (holistic processing) and can statistically summarize facial expressions in a crowd (ensemble coding). Previously, we tested if the two kinds of processing relied on the same mechanisms (Sun & Chong, 2016). We utilized the face inversion effect to disrupt emotion judgments of a single and multiple faces, and found that the face inversion effect was weaker with multiple faces than a single face, suggesting that the holistic processing of individual faces and the ensemble coding of multiple faces are rooted in different mechanisms. We hypothesized that interferences generated by the inversion could have differed in assessing emotions of a single and multiple faces. As the mouth carries more information than the eyes in a happiness judgment (Calvo et al., 2014), facial features (the left and right eyes, nose, and mouth), unlike individual faces in a crowd, could have unevenly contributed to the judgments of specific emotions. Thus, we varied the number of inverted features in single-face judgments and that of inverted faces in multiple-face judgments. Specifically, four features within a face or four faces shown in one side from fixation could be all upright, all inverted, or two upright and two inverted. Participants judged the emotional intensity of a single face or the average intensity of the four faces compared to a single face shown on the other side. Results showed that the intensity discrimination became poor as more inverted features and faces were shown. Moreover, whereas happy judgments with a single face were disrupted mostly by the mouth inversion, the average intensity judgments were equally influenced by all faces regardless of their orientation. These results suggest that the different inversion effects arose from differential contributions of components to the perception of single or mean facial expression.
Meeting abstract presented at VSS 2017