The Shared Signal Hypothesis in the Visual Search Paradigm with Real Faces

Abstract

According to the shared signal hypothesis, the perception of faces is enhanced when the emotion, gaze, and motivation are congruent. When threatening emotions are congruent with gaze and motivation, it becomes relevant to the observer. For example, an angry face looking at you is a direct threat whereas a fearful face looking away signals threat in the environment relevant to you. Visual search studies using faces have yielded mixed results and most studies have been compromised by low ecological validity. This study aims to tackle this limitation by using real faces. 32 non-clinical participants completed a visual search task with multiple trials representing each combination between the levels of emotion (angry, fearful), gaze (direct, averted) and set-size (4, 8, 16). Gaze data were retrieved with an eye-tracker and traitanxiety was measured with a questionnaire afterwards. Multilevel models were performed to assess differences in response times between all possible combinations. The results showed that the shared signal hypothesis was only true for anger. Furthermore, fearful faces were found faster than angry faces and once found, they were also faster identified as the emotional target. Lastly, trait-anxiety levels did not moderate reaction times for self-relevant threat, but did bias the individual to direct gazes compared to indirect gazes. Limitations include the lack of an emotional intensity measurement of the stimuli and the small sample of models used for the stimuli. The current study is a great stepping stone for future research investigating the ecological validity of attentional biases to self-relevant facial threat.