Abstract
It is still unclear how informative peripheral visual information is to emotional face perception and whether different types of facial expressions are similarly affected by the distance from central vision (eccentricity). Here we examined how eccentricity influences perception of facial expressions’ emotional valence and whether different emotional valences are differentially affected. Images of facial expressions with positive (pleasant, happy), negative (unpleasant, afraid), or neutral valence were presented in 10s blocks where each block was of a specific eccentricity (0°, 2°, or 4°) and consisted of a dominant emotional valence (8 images, 200ms/image, 1050ms ISI). Participants (n=51) were instructed to keep fixation and report the dominant valence across each block while their eye movements were being monitored. As expected, eccentricity significantly influenced facial emotional valence perception such that accuracy decreased and RTs increased with growing eccentricity. Additionally, we found that negative valence accuracy decreased with eccentricity significantly more than that of positive valence. As pupil size has been suggested as a measure of emotional arousal, we examined whether emotional arousal, as measured by pupil dilation across the 10s blocks, could potentially explain the valence and/or the eccentricity effects. However neither of these factors modulated pupillary changes suggesting that the perceptual effects we found were not likely to be related to arousal (i.e. increased sympathetic activity). Our parafoveal behavioral investigations indicate that the mechanisms supporting perception of different facial valence may be dissociated and suggest that peripheral visual information may further our understanding of the mechanisms underlying perception of facial expressions of emotions.