Asymmetries During Multiple Face Encoding: Increased Dwell Time and Number of Fixations in the Upper Visual Hemifield.

Fatima M Felisberti,Liam Currie

doi:10.1177/2041669519827974

Abstract

Visual field asymmetries in the encoding of groups of faces have rarely been investigated. Here, eye movements (percentage of dwell time [pDT] and number of fixations [nFix]) were recorded during the encoding of three groups of four faces tagged with cheating, cooperative, or neutral behaviours. Faces in each group were placed in the top left, top right, bottom left, or bottom right quadrants. Face recall was equally high in the three behavioural groups. Conversely, pDT and nFix were higher for faces in the upper hemifields. Most of the first saccades were made to the top left visual quadrant, which also commanded a higher pDT and nFix than the other quadrants. The findings are relevant to the understanding of visual field asymmetries in the processing of multiple faces, a common social scenario, and may be linked to reading habits in conjunction (or not) with cultural and environmental cues.

Highlights

Face-to-face communication is at the core of human social life and it relies strongly on our ability to recognize the faces of people we have interacted with, to develop stable future associations and exchanges (Rand, Dreber, Ellingsen, Fudenberg, & Nowak, 2009) and to avoid conflict or financial losses (Blais, Jack, Scheepers, Fiset, & Caldara, 2008; Bonner, Burton, & Bruce, 2003)
This study examined if there were perceptual visual field asymmetries (VFAs) in eye movements during the encoding of multiple faces, as suggested by the different recognition accuracy for faces presented at different locations in the visual field (Felisberti & McDermott, 2013)
VFAs in face processing have been reported in studies using a wide range of experimental paradigms, and many showed an advantage of the upper visual field for object recognition and faces

Summary

Introduction

Face-to-face communication is at the core of human social life and it relies strongly on our ability to recognize the faces of people we have interacted with, to develop stable future associations and exchanges (Rand, Dreber, Ellingsen, Fudenberg, & Nowak, 2009) and to avoid conflict or financial losses (Blais, Jack, Scheepers, Fiset, & Caldara, 2008; Bonner, Burton, & Bruce, 2003). A well-developed ability to recognize faces linked to distinct types of behaviour is evolutionarily advantageous. Few studies examined face recognition with multiple faces at once and even fewer investigated tagging faces with different behaviours, a common behaviour in social gatherings and relevant information in eye witnesses’ reports. The findings from earlier studies showed a range of face recognition biases in social scenarios, from biases towards cooperators (cf., Barclay, 2004) or free-riders (cf., Cosmides, Tooby, Fiddick, & Bryant, 2005), to reports of no biases towards either of them (cf., Felisberti & Farrelly, 2016). One such study revealed recognition biases towards faces tagged with a cooperative rather than neutral or cheating behaviour (Felisberti & Pavey, 2010). The question of whether we have any a priori bias is still open

Methods

Results

Conclusion