Functional architecture of visual emotion recognition ability: a latent variable approach

Emotion recognition has been a focus of considerable attention for several decades. However, despite this interest, the underlying structure of individual differences in emotion recognition ability has been largely overlooked and thus is poorly understood. For example, limited knowledge exists concerning whether recognition ability for one emotion (e.g., disgust) generalizes to other emotions (e.g., anger, fear). Furthermore, it is unclear whether emotion recognition ability generalizes across modalities, such that those who are good at recognizing emotions from the face, for example, are also good at identifying emotions from nonfacial cues (such as cues conveyed via the body). The primary goal of the current set of studies was to address these questions through establishing the structure of individual differences in visual emotion recognition ability. In three independent samples (Study 1: n=640; Study 2: n=389; Study 3: n=303), we observed that the ability to recognize visually presented emotions is based on different sources of variation: a supramodal Emotion-General factor, supramodal Emotion-Specific factors, and Face-And Within-Modality Emotion-Specific factors. In addition, we found evidence that general intelligence and alexithymia were associated with supramodal emotion recognition ability. Autism-Like traits, empathic concern, and alexithymia were independently associated with Face-Specific emotion recognition ability. These results (a) provide a platform for further individual differences research on emotion recognition ability, (b) indicate that differentiating levels within the architecture of emotion recognition ability is of high importance, and (c) show that the capacity to understand expressions of emotion in others is linked to broader affective and cognitive processes.