Abstract

A temporal two-interval forced-choice paradigm is used to evaluate the relative strength of the visual signals conveyed by the eyes and the mouth in happy and fearful facial expressions.  Stimuli were black and white images of faces with a neutral, happy or fearful expression.  The happy and fearful visual signals were conveyed by the eyes, the mouth or by the whole face. A range of signal strengths (0-100%) were created by morphing the neutral and expressive images.  One interval contained the neutral face (0%) and the other the expressive face (varied from 0 – 100%, presented for 200ms).  Observers indicated the interval with the more expressive image.  Performance increased from chance (50%) to 100% correct as signal strength increased in all conditions.

Psychometric functions for happy expressions were shifted to the left of those for fearful expressions indicating that observers are more sensitive to happy expressions.  This suggests that the emotion signals conveyed by a happy face are more salient than those conveyed by a fearful face.

For happy expressions, psychometric functions for conditions with an expressive mouth were shifted to the left of those in which only the eyes were expressive. For fearful expressions, psychometric functions for conditions in which the eyes were expressive were shifted to the left of those in which only the mouth was expressive. This suggests that the visual signals conveyed by the mouth are more salient for happy facial expressions and the visual signals conveyed by the eyes are more salient for fearful expressions.

Future research is aimed at understanding the extent to which this perceptual difference can be explained by the nature of the low level visual signal or higher cognitive function.