Recognition thresholds for static and dynamic emotional faces

Abstract

We investigated the minimum expressive intensity that is required to recognize (above chance) static and dynamic facial expressions of happiness, sadness, anger, disgust, fear, and surprise. To this end, we varied the degree of intensity of emotional expressions unfolding from a neutral face, by means of graphics morphing software. The resulting face stimuli (photographs and short videos) were presented in an expression categorization task for 1 s each, and measures of sensitivity or discrimination (A') were collected to establish thresholds. A number of physical, perceptual, categorical, and affective controls were performed. All six basic emotions were reliably recognized above chance level from low intensities, although recognition thresholds varied for different expressions: 20% of intensity, for happiness; 40%, for sadness, surprise, anger, and disgust; and 50%, for fear. The advantage of happy faces may be due to their greater physical change in facial features (as shown by automated facial expression measurement), also at low levels of intensity, relative to neutral faces. Recognition thresholds and the pattern of confusions across expressions were, nevertheless, equivalent for dynamic and static expressions, although dynamic expressions were recognized more accurately and faster.