Purchase this article with an account.
Michael D. Anes, Holly K. Sprunger, Mark P. Heilala; Identification performance of brief dynamic emotional expressions as a function of orientation and position in the visual periphery. Journal of Vision 2002;2(7):594. doi: 10.1167/2.7.594.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent work indicates that negative emotional expressions are perceived outside the current locus of attention in visual search displays (Eastwood et al., 2001). Other research has found powerful behavioral and physiological effects of viewing brief, masked anger and fear expressions. We try here to tie together this work with recent facial motion research: motion is an independent information source in facial identification, and humans are quite sensitive to the normal temporal characteristics of emotional expressions.
Participants identified the displayed emotional expression of brief (330 ms) 6-frame movies of three Emotions (anger, surprise, happiness) and of static versions of the same stimuli. Stimuli were displayed in windows subtending 1.4°, either in the center of the screen, or offset 2.1° or 4.2° to the left and right of center, in upright and inverted orientations. Expected negative effects of Inversion (F[1,15]=12.91, p<.01) were found, as were effects of Emotion (identification accuracy for happiness and surprise were > 90%, anger displays ∼84%, F[2,30]=13.89, p<.001) and Position (with expressions recognized more accurately in the center and near periphery than more peripherally, F[2,30]=9.41, p<.01). Significant two- and three-way interactions occurred in the context of a significant four-way interaction of the factors, (F[4,60]=3.86, p<.01). There was no indication that anger was identified in the periphery better than other emotions, either with or without motion; however upright and inverted dynamic anger displays showed the greatest benefit in central compared to peripheral positions. In contrast, static, upright happiness displays showed an identification accuracy decrease centrally, due to a speed-accuracy tradeoff. An additional experiment is underway with a 200-ms display duration coupled with the ability to eliminate trials in which eye movements occurred, and in which we explicitly investigate sex differences in performance.
Sponsor: Foundation for Undergraduate Research, Scholarship and Creative Activity (FURSCA), Albion College
This PDF is available to Subscribers Only