In 2 experiments, participants were presented schematic faces with emotional expressions (threatening, friendly) in a neutral-faces context or neutral expressions in an emotional-faces context. These conditions were compared with detection performance in displays containing key features of emotional faces not forming the perceptual gestalt of a face. Supporting the notion of a threat detection advantage, Experiment 1 found that threatening faces were faster detected than friendly faces, whereas no difference emerged between the corresponding feature conditions. Experiment 2 increased task difficulty with a backward masking procedure and found corresponding results. In neither of the studies was the threat detection advantage associated with reduced accuracy. However, features were, in general, detected faster than faces when task difficulty was high.
2006 APA, all rights reserved