Elsevier

Brain and Cognition

Volume 67, Issue 2, July 2008, Pages 127-139
Brain and Cognition

Is anyone looking at me? Direct gaze detection in children with and without autism

https://doi.org/10.1016/j.bandc.2007.12.001Get rights and content

Abstract

Atypical processing of eye contact is one of the significant characteristics of individuals with autism, but the mechanism underlying atypical direct gaze processing is still unclear. This study used a visual search paradigm to examine whether the facial context would affect direct gaze detection in children with autism. Participants were asked to detect target gazes presented among distracters with different gaze directions. The target gazes were either direct gaze or averted gaze, which were either presented alone (Experiment 1) or within facial context (Experiment 2). As with the typically developing children, the children with autism, were faster and more efficient to detect direct gaze than averted gaze, whether or not the eyes were presented alone or within faces. In addition, face inversion distorted efficient direct gaze detection in typically developing children, but not in children with autism. These results suggest that children with autism use featural information to detect direct gaze, whereas typically developing children use configural information to detect direct gaze.

Introduction

Information gained from another person’s eyes plays a crucial role in human social communication. Among various functions of gaze processing, detection of direct gaze or eye contact is essential in social interaction and communication. Direct gaze signals the intention of the gazer towards the perceiver. Eye contact also plays a major role in communication and affective bonding (Kleinke, 1986, Robson, 1967, Robson et al., 1969). Csibra and Gergely (2006) argue that perceived eye contact signals communicative ostention, and initiates referential communication.

Experimental studies have found that direct gaze affects perception, cognition and attention. For example, in visual search, target faces with direct eye gaze are detected faster and more efficiently than those with averted eye gaze (Conty et al., 2006, von Grünau and Anston, 1995, Senju et al., 2005a; Senju & Hasegawa, 2006). In addition, when the gaze direction of others is ambiguous and difficult to perceive, people are biased to judge the gaze as “looking at me” (Martin and Jones, 1982, Martin and Rovira, 1981, Martin and Rovira, 1982). Direct gaze also holds attention and makes it difficult to disengage from the face (Senju & Hasegawa, 2005). In addition, faces with direct gaze were remembered better than faces with averted gaze (Hood et al., 2003, Mason et al., 2004, Smith et al., 2006, Vuilleumier et al., 2005). It is also known that a stranger gazing directly at the perceiver increases autonomic arousal in adults (Gale et al., 1978, Gale et al., 1975, Nichols and Champness, 1971).

Failure to develop typical mutual gaze behavior is one of the core symptoms of severe social and communicative disorders, and of autism (American Psychiatric Association, 1994, Baron-Cohen, 1995). Retrospective home video analyses found that from the first year of life, infants who were later diagnosed with Autism Spectrum Disorders (ASD) orient less to faces than typically developing infants (Baranek, 1999, Clifford et al., 2007, Maestro et al., 2005, Osterling and Dawson, 1994, Osterling et al., 2002, Werner and Dawson, 2005). Hobson and Lee (1998) also reported that older children and adolescents with ASD make eye contact less in a communicative context (greeting) than those without ASD. Studies with eye-tracking techniques confirm these observations and revealed that individuals with ASD fixate less to eyes compared to typically developing individuals (Dalton et al., 2005, Klin et al., 2002, Neumann et al., 2006, Pelphrey et al., 2002, Spezio et al., 2007, but see also van der Geest, Kemner, Verbaten, & van Engeland, 2002).

Although these observation studies are very informative for spontaneous behaviour, they do not clarify how individuals with ASD process direct gaze, or whether perceived direct gaze affects cognition in individuals with ASD. Moreover, there are few studies which have empirically examined the cognitive and neural basis of eye contact processing in ASD. Furthermore, of these experimental studies that investigate eye contact processing in ASD, the findings are inconsistent. A series of experimental studies byour group found that individuals with ASD failed to show the facilitated behavioural (Senju, Yaguchi, Tojo, & Hasegawa, 2003) and event-related potential (ERP) (Senju, Tojo, Yaguchi, & Hasegawa, 2005b) responses associated with direct gaze. On the other hand, other neurophysiological studies reported that individuals with ASD elicited large ERP or magnetoencephalography signals in response to direct gaze, whereas this was not apparent in typically developing individuals (Grice et al., 2005, Kylliäinen et al., 2006). In addition, Kylliäinen and Hietanen (2006) presented looming faces with direct or averted gaze, and found that looming faces with either gaze direction, elicited a similar skin conductance response (SCR) in typically developing individuals. However, individuals with ASD, elicited a larger SCR in response to a looming face with direct gaze than one with averted gaze. It is difficult to interpret the cognitive and/or affective basis of the SCR response because the looming feature of the stimuli differed from other studies, and because the SCR response was smaller in individuals with ASD compared to typically developing individuals. However, at least, the differential response to gaze suggests that individuals with ASD possess a sensitivity to others’ direct gaze.

Interestingly, one of our previous studies (Senju, Hasegawa, & Tojo, 2005a) found conflicting results about direct gaze detection in autism. This study adopted a visual search paradigm initially used by von Grünau and Anston (1995), in which eye stimuli with various gaze directions were presented. Participants were instructed to detect targets of a particular eye direction, i.e. direct gaze, within a set of distracters of a different eye direction, i.e. averted gaze (Fig. 1). There were two versions of the task, in the first we used schematic eyes (Fig. 1a) as used by von Grünau and Anston (1995), and in the second we used photographs (Fig. 1b). In the first experiment, children with autism, as well as typically developing children, showed the ‘stare-in-the-crowd’ effect (or asymmetry in search performance), performing better for the detection of direct gaze than the detection of averted gaze. In contrast, when the gazes were presented in photographs of laterally oriented faces, typically developing children were faster to detect direct gaze than averted gaze, but gaze direction did not affect search performance in children with autism. In addition, the faster detection of direct gaze in typically developing children was limited within the context of an upright face, when given inverted face stimuli their search performance between different gazes was no longer significantly different.

There are several possibilities why children with ASD were faster to detect direct gaze in schematic eyes but not in laterally oriented photographic faces. Firstly, as Conty and colleagues argued (Conty et al., 2006), features specific to the schematic stimuli such as an unrealistically close distance between the two eyes or the high contrast of the eyes against the white background may have helped children with autism to detect direct gaze. Secondly, the presence of a whole face may have distracted children with autism and interfered with their attention. For example, while observing faces, individuals with autism fixate less to the eye region than typically developing individuals (Dalton et al., 2005, Klin et al., 2002, Neumann et al., 2006, Pelphrey et al., 2002, Spezio et al., 2007, but see also van der Geest et al., 2002). Finally, they may have failed to integrate the eye direction and the facial orientation to detect the gaze direction in the visual search task. Although individuals with ASD show the face inversion effect, they can recognize a face when presented with parts of a face just as well as when presented with a whole face. The results contrasted with typically developing individuals who benefited from the presence of whole face (Joseph and Tanaka, 2003, Teunisse and de Gelder, 2003). Since other studies have shown the presence of configural face processing capacities in individuals with ASD (Lahaie et al., 2006, Rouse et al., 2004), these results may not suggest an impairment in configural face processing, but rather the presence of a cognitive style that spontaneously prefers featural processing and/or enhances an individual’s perceptual ability to process details. Therefore, in the aforementioned study (Senju et al., 2005a) one needs to perceive eye direction within the context of facial orientation (Wollaston, 1824) in order to detect direct gaze in laterally oriented face. Thus, it is possible that the individuals with ASD fail to show faster detection for direct gaze when the gaze can only be perceived with reference to facial orientation.

In the current paper, we used a visual search paradigm to examine the three hypotheses outlined above. Typically in a visual search paradigm, a target is presented within a varying number of distracters, and one’s task is to detect the target as fast as possible (e.g. Treisman and Souther, 1985, Wolfe, 2001). Previous studies demonstrated that individuals with ASD are faster at a visual search task than typically developing individuals when non-social stimuli were used as targets and distracters (O’Riordan, 2004, O’Riordan and Plaisted, 2001, O’Riordan et al., 2001, Plaisted et al., 1998). They also show a shallower ‘search slope’ (the search latency divided by the number of distracters), suggesting that their search performances are more efficient. It is advantageous to use such tasks because any atypical performance in individuals with ASD cannot be attributed to the general difficulty of the task involved. To date, at least two studies have used social stimuli in a visual search task with individuals with ASD (Ashwin et al., 2006, Senju et al., 2005a). Both studies failed to find superior visual search performance in the ASD group, but found that their speed and accuracy equaled that of typically developing individuals.

The current experiments used the same design as Senju et al., (2005a; Experiment 2) but with different stimuli. The stimuli in Experiment 1 were static images of the eye region taken from forward-facing photographs of faces (Fig. 2). The aim of Experiment 1 was to test the first hypothesis that specific features of schematic stimuli used in Senju et al. (2005a) helped children with autism to detect direct gaze. If the faster direct gaze detection found in children with ASD is limited to the specific information contained in schematic eyes used in Senju et al. (2005a), the search performance of children with autism would not be modulated by the gaze direction in response to photographic eyes. Experiment 2 used images of front-view faces to test the second hypothesis that the presence of the face would interfere with direct gaze detection in children with autism. If the mere presence of facial context interrupts attention of children with autism, they should show the ‘stare-in-the-crowd’ effect in Experiment 1, but not in Experiment 2. In contrast, if the difficulty in direct gaze detection in children with autism is limited to the integration of face and eye directions, they should be faster to detect direct gaze than averted gaze both in Experiments 1 and 2. As a result, when the isolated eyes were presented (Experiment1), children with ASD were faster to detect direct gaze than averted gaze, just like typically developing children. When the whole faces were presented, typically developing children demonstrated faster direct gaze detection only when the faces were presented right side up, but children with ASD were faster to detect direct gaze regardless of the facial orientation (upright or inverted). These results suggest that children with ASD rely on featural information to detect direct gaze, whereas typically developing children use configural information to detect direct gaze.

Section snippets

Experiment 1

This experiment was designed to investigate whether children with autism show the ‘stare-in-the-crowd’ effect when the real images of eye regions were used as stimuli. Stimuli were images of eye regions with various gaze directions, which were cropped from photographs of the same female face (see Fig. 2). Targets were either direct gaze or averted gaze. If children with autism have sensitivity to realistic direct gaze, they should be faster and more efficient to detect targets with direct gaze

Experiment 2

In Experiment 2, we used whole, front-view faces (Fig. 3) as stimuli to investigate whether the presence of facial context would impede the ‘stare-in-the-crowd’ effect in children with autism. As we described before, children with autism do not show the ‘stare-in-the-crowd’ effect when the eyes were presented in the context of laterally oriented faces (Senju et al., 2005a, Experiment 2, see also Fig. 1b). Because Experiment 1 confirmed that children with autism, as well as typically developing

General discussion

The current study utilized a visual search paradigm to investigate direct gaze detection in children with autism. The findings clearly demonstrate that children with autism detect direct gaze faster and more efficiently than averted gaze, regardless of whether the eyes were presented alone (Experiment 1) or within a facial context (Experiment 2). These results suggest that the direct gaze, at least within front-view faces, is salient for individuals with autism as well as for typically

Conclusion

The current study used a visual search paradigm to investigate whether direct gaze in front-view faces facilitate detection of the target in children with and without autism. In Experiment 1, when isolated eyes were presented, both typically developing children and children with autism were faster to detect direct gaze compared to averted gaze, and no group differences were observed. In Experiment 2, when the images of whole, front-view faces were used as stimuli, typically developing children

Acknowledgments

We thank all the participants and parents who supported our study. We acknowledge Yura Kunihira and Hironori Akechi for help in data collection, Gergely Csibra, Laurence Conty, Mayada Elsabbagh, Nathalie George, Mark H Johnson and Sarah Lloyd-Fox for discussions and comments on earlier version of the draft, Kikue Sakaguchi for allowing us to use her photographs as stimuli in Experiment 1, and Bruce Hood for allowing us to use the stimuli in Experiment 2. This study was supported by the Japan

References (74)

  • A. Senju et al.

    Deviant gaze processing in children with autism: An ERP study

    Neuropsychologia

    (2005)
  • J.P. Teunisse et al.

    Face processing in adolescents with autistic disorder: The inversion and composite effects

    Brain and Cognition

    (2003)
  • American Psychiatric Association, (1994). Diagnostic and statistical manual of mental disorders (4th ed.). DSM-IV....
  • G.T. Baranek

    Autism during infancy: A retrospective video analysis of sensory-motor and social behaviors at 9–12 months of age

    Journal of Autism and Developmental Disorders

    (1999)
  • S. Baron-Cohen

    Mindblindness: An essay on autism and theory of mind

    (1995)
  • S. Clifford et al.

    Assessing the early characteristics of autistic disorder using video analysis

    Journal of Autism and Developmental Disorders

    (2007)
  • L. Conty et al.

    Searching for asymmetries in the detection of gaze contact versus averted gaze under different head views: A behavioural study

    Spatial Vision

    (2006)
  • G. Csibra et al.

    Social learning and social cognition: The case for pedagogy

  • K.M. Dalton et al.

    Gaze fixation and the neural circuitry of face processing in autism

    Nature Neuroscience

    (2005)
  • F. Happé et al.

    The weak coherence account: Detail-focused cognitive style in autism spectrum disorders

    Journal of Autism and Developmental Disorders

    (2006)
  • R.P. Hobson et al.

    Hello and goodbye: A study of social engagement in autism

    Journal of Autism and Developmental Disorders

    (1998)
  • B.M. Hood et al.

    Eye remember you: The effects of gaze direction on face recognition in children and adults

    Developmental Science

    (2003)
  • Japanese WISC-III Publication Committee. (1998). Nihonban WISC-III chinou kensahou (Japanese Wechsler Intelligence...
  • J. Jenkins et al.

    Configural processing in the perception of eye-gaze direction

    Perception

    (2003)
  • M.H. Johnson et al.

    The emergence of the social brain network: Evidence from typical and atypical development

    Development and Psychopathology

    (2005)
  • R.M. Joseph et al.

    Holistic and part-based face recognition in children with autism

    Journal of Child Psychology and Psychiatry

    (2003)
  • C. Kleinke

    Gaze and eye-contact: A research review

    Psychological Bulletin

    (1986)
  • A. Klin et al.

    Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism

    Archives of General Psychiatry

    (2002)
  • A. Kylliäinen et al.

    Face- and gaze-sensitive neural responses in children with autism: A magnetoencephalographic study

    European Journal of Neuroscience

    (2006)
  • A. Kylliäinen et al.

    Skin conductance responses to another person’s gaze in children with autism

    Journal of Autism and Developmental Disorders

    (2006)
  • A. Lahaie et al.

    Face perception in high-functioning autistic adults: Evidence for superior processing of face parts, not for a configural face-processing deficit

    Neuropsychology

    (2006)
  • S.R. Langton

    The mutual influence of gaze and head orientation in the analysis of social attention direction

    Quarterly Journal of Experimental Psychology

    (2000)
  • B. Lopez et al.

    Face processing in high-functioning adolescents with autism: Evidence for weak central coherence

    Visual Cognition

    (2004)
  • S. Maestro et al.

    How young children treat objects and people: An empirical study of the first year of life in autism

    Child Psychiatry and Human Development

    (2005)
  • W.W. Martin et al.

    The accuracy of eye-gaze judgment—A signal-detection approach

    British Journal of Social Psychology

    (1982)
  • W.W. Martin et al.

    An experimental-analysis of discriminability and bias in eye-gaze judgment

    Journal of Nonverbal Behavior

    (1981)
  • W.W. Martin et al.

    Response biases in eye-gaze perception

    Journal of Psychology

    (1982)
  • Cited by (78)

    • Task switching in autism: An EEG study on intentions and actions

      2018, Neuropsychologia
      Citation Excerpt :

      Indications for the presence of such delayed adjustments to the task requirements are observed in the behavioural data, in terms of increased switch costs in the ASC group. Interestingly, in the domain of attentional orienting, it has already been observed that children with ASC have an increased reliance on bottom-up strategies (Amso et al., 2014; Senju,Senju et al., 2008), and that a possible explanation for some of these results may be that people with ASC have difficulties with developing top-down strategies for attending to, for example, socially relevant stimuli (Neumann et al., 2006). As such, inadequate top-down control in ASC should be taken into account as a potential explanation for the observed impairments in both the domains of social interactions and task switching.

    • A strategic plan to identify key neurophysiological mechanisms and brain circuits in autism

      2018, Journal of Chemical Neuroanatomy
      Citation Excerpt :

      We therefore suggest that while atypicalities in sensory and perceptual processing are not specific to the social domain (Clery et al., 2013c; Kovarski et al., 2017), they could strongly affect the development of social and emotional abilities in children with ASD. Atypicalities in face and biological movement processing have also been noted in the domain of gaze perception (Senju et al., 2003, 2004, 2005, 2008), particularly in infants where it seems to be potentially predictive of autism emergence (Elsabbagh et al., 2012). Beside facial expressions, social interactions are also supported by visuo-posturo-motor adjustments.

    View all citing articles on Scopus
    View full text