Attentional biases to faces with direct versus averted gaze in children without and with autism spectrum disorder: A dot-probe paradigm

Abstract

Previous research utilized a variety of paradigms to demonstrate attentional biases to faces with direct versus averted gaze in typical development (TD) and no such bias in autism spectrum disorder (ASD). However, little is known about whether the biases can be observed during automatic processing stages in TD and whether the lack of such bias in ASD is due to passive neglect or active avoidance of direct gaze. Therefore, we employed a dot-probe paradigm and manipulated stimulus onset asynchrony (SOA) to measure response times to probes replacing faces with direct gaze compared to faces with averted gaze in children without and with ASD. There was no evidence of attentional bias in either group when stimuli were presented during the automatic processing stage (200 ms SOA). However, during the controlled processing stage (1000 ms SOA), an attentional bias to faces with direct versus averted gaze was found in control children but not in those with ASD. The results indicate that the facilitation of direct gaze on facial attention occurs during controlled rather than automatic processing stages in TD individuals. In contrast, children with ASD respond indiscriminately to direct and averted gaze during both stages of cognitive processing, supporting the gaze indifference hypothesis. For TD children but not for children with ASD, direct gaze is an adaptively informative or socially salient signal.