posted on 2020-06-22, 12:26authored byMinke J. de Boer, Deniz Başkent, Frans W. Cornelissen
<p>The majority of emotional
expressions used in daily communication are multimodal and dynamic in nature.
Consequently, one would expect that human observers utilize specific perceptual
strategies to process emotions and to handle the multimodal and dynamic nature
of emotions. However, our present knowledge on these strategies is scarce,
primarily because most studies on emotion perception have not fully covered
this variation, and instead used static and/or unimodal stimuli with few
emotion categories. To resolve this knowledge gap, the present study examined how
dynamic emotional auditory and visual information is integrated into a unified
percept. Since there is a broad spectrum of possible forms of integration, both
eye movements and accuracy of emotion identification were evaluated while
observers performed an emotion identification task in one of three conditions:
audio-only, visual-only video, or audiovisual video. In terms of adaptations of
perceptual strategies, eye movement results showed a shift in fixations toward the
eyes and away from the nose and mouth when audio is added. Notably, in terms of
task performance, audio-only performance was mostly significantly worse than
video-only and audiovisual performances, but performance in the latter two
conditions was often not different. These results suggest that individuals
flexibly and momentarily adapt their perceptual strategies to changes in the
available information for emotion recognition, and these changes can be comprehensively
quantified with eye tracking.</p>