Attentional capture by fearful faces requires consciousness and is modulated by task-relevancy: a dot-probe EEG study

In the current EEG study, we used a dot-probe task in conjunction with backward masking to examine the neural activity underlying awareness and spatial processing of fearful faces and the neural processes for subsequent cued spatial targets. We presented face images under different viewing conditions (subliminal and supraliminal) and manipulated the relation between a fearful face in the pair and a subsequent target. Through both mass univariate analysis and multivariate pattern analysis, we found that fearful faces can be processed to an extent where they attract spatial attention only when they are presented supraliminally and when they are task-relevant. The spatial attention capture by fearful faces also modulated the processing of subsequent lateralised targets that were spatially congruent with the fearful face, in both behavioural and neural data. There was no evidence for nonconscious processing of the fearful faces in the current paradigm. We conclude that spatial attentional capture by fearful faces requires visual awareness and it is modulated by top-down task demands.