Dissociating Affective and Perceptual Effects of Schematic Faces on Attentional Scope

Attention allocation to positive and negative stimuli differs. For example, the flanker-interference asymmetry describes a pattern of results on flanker tasks using emotional stimuli, where a typical flanker-interference effect is observed for positive targets but not for negative targets. There are two dominant explanations for the flanker-interference asymmetry. According to the emotion-first explanation, negative targets are preferentially processed to facilitate the processing of potentially threatening stimuli. In contrast, feature-first explanations argue that the asymmetry results from differences in perceptual complexity between positive and negative stimuli. Three experiments used schematic emotional faces in a flanker task to directly compare these explanations. To manipulate the perceptual complexity of the stimuli, an enclosing circle was present on half of the trials. In all three experiments, reaction times showed the expected flanker-interference asymmetry, but the pattern was not influenced by the presence of the circle. However, event-related potentials showed that perceptual complexity influenced both the structural encoding and evaluative processing of the faces in the N170 and P3b time windows. These results suggest that both perceptual complexity and emotional valence play an important role in the processing of schematic emotional faces, but that emotional valence may have a stronger effect at evaluative stages of processing. Other findings show that the enclosing circle may alter the perceived emotional expression of neutral faces.Public Significance StatementAttention to emotional stimuli plays a critical role in everyday functioning, yet it remains unclear if affective processing is distinct from the processing of perceptual features. This study suggests that these two processes are fully dissociable; that is, they are important at different stages of stimulus processing and have distinct effects on responses.