SEARCH

SEARCH BY CITATION

Keywords:

  • autism spectrum disorder;
  • facial expression;
  • local processing;
  • spatial frequencies;
  • inversion effect

Abstract

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

To investigate whether facial expression processing in children with autism spectrum disorder (ASD) is based on local information of the stimuli, we prepared low spatial frequency (LSF) images with blurred facial features and high spatial frequency (HSF) images with rich facial features from broad (normal) spatial frequency (BSF) images. Eighteen children with ASD (mean age 11.9 years) and 19 typically developing (TD) children (mean age 11.4 years) matched on nonverbal IQ were presented these stimuli in upright and inverted orientations. The children with ASD had difficulty in processing facial expressions from the BSF and LSF images, but not from the HSF images. In addition, the BSF and HSF images elicited the inversion effect in the TD children, but not in the children with ASD. In contrast, the LSF images elicited the inversion effect in both groups of children. These results suggest that children with ASD are biased towards processing facial expression based on local information, even though their capacity to process facial expressions configurally is spared.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by difficulty in communicating effectively with other people and developing social relationships, and by restricted, repetitive, and stereotyped patterns of behavior, interests, and activities (American Psychological Association (APA), 1994). ASD is also characterized by a cognitive style biased toward local rather than global information processing, which is termed weak central coherence (WCC) (Happé, 1999; Happé & Frith, 2006). For example, children with ASD performed better than controls on an embedded-figures task (Pellicano, Gibson, Maybery, Durkin, & Badcock, 2005; Shah & Frith, 1983, but see also White & Saldaña, 2011). Plaisted, Swettenham, and Rees (1999) found a local advantage on the Navon task in children with ASD. Moreover, because individuals with ASD showed this detail-focused perception with intact global processing (Mottron, Burack, Iarocci, Belleville, & Enns, 2003), the enhanced perceptual functioning (EPF) model was proposed (Mottron, Dawson, Soulières, Hubert, & Burack, 2006). Local-based processing in children with ASD has been found in face perception as well as nonface perception, which suggested a general perceptual deficit in children with ASD (Davies, Bishop, Manstead, & Tantam, 1994).

Face processing in children with ASD

Mild to moderate deficits in face discrimination and recognition have been reported in children with ASD (Gliga & Csibra, 2007). Children with ASD perform less accurately than controls in face recognition (Klin, Sparrow, de Bildt, Cicchetti, Cohe, & Volkmar, 1999; Robel, Ennouri, Piana, Vaivre-Douret, Perier, Flament, & Mouren-Simeoni, 2004), even when the faces are familiar (Boucher, Lewis, & Collis, 1998). However, several studies have failed to find deficits in face recognition in individuals with ASD (Celani, Battacchi, & Arcidiacono, 1999; Deruelle, Rondan, Gepner, & Tardif, 2004; Gepner, Deruelle, & Grynfeltt, 2001; Volkmar, Sparrow, Rende, & Cohen, 1989). Sasson (2006) proposed that the types of experimental tasks, the participant age, and their developmental level (i.e., “classic” or high-functioning autism) may contribute to this discrepancy. Furthermore, as children with ASD grow older, they may develop compensatory strategies for processing faces, which could help minimize their general impairment (Klin et al., 1999). In this regard, it has been reported that individuals with ASD tend to process faces in an atypical way. They seem to have a more feature-based rather than configuration-based processing style for faces.

It appears that, compared with other stimuli, faces are disproportionately sensitive to inversion (Yin, 1969), and this difference in performance between upright and inverted faces is called the inversion effect. The inversion effect demonstrates that faces are processed as an overall template in a holistic manner and that the spatial relations between facial features are processed configurally. Inversion disrupts this holistic and configural processing, and inverted faces are then processed in a piecemeal fashion like objects. However, individuals with ASD performed well on tasks with inverted faces and showed only a small inversion effect. Langdell (1978) found that children with ASD were significantly better than controls at recognizing their peers in the inverted mode. Inversion of a face did not interfere with the ability to discriminate faces in children (Rose, Lincoln, Lai, Ene, Searcy, & Bellugi, 2007) and adolescents (Hobson, Ouston, & Lee, 1988) with ASD relative to controls. Although high-functioning adolescents (Teunisse & de Gelder, 2003) and adults (Lahaie, Mottron, Arguin, Berthiaume, Jemel, & Saumier, 2006; Nishimura, Rutherford, & Maurer, 2008) with ASD demonstrated a face inversion effect, Faja, Webb, Merkle, Aylward, and Dawson (2009) suggested that the inversion effect was weakened in individuals with ASD, based on the data that the difference in reaction time for upright versus inverted faces was only half as large in adults with ASD than in controls. This smaller inversion effect in individuals with ASD is taken as a sign of local rather than configural processing of faces because configural information is less available, and only local information is available when a face is presented upside-down.

An atypical style of face processing in individuals with ASD has been demonstrated by another technique that exploits the composite effect (Young, Hellawell, & Hay, 1987). Teunisse and de Gelder (2003) failed to observe a normal composite effect in adolescents with ASD (but see Nishimura et al., 2008, for contrary evidence). In a study by Young et al. (1987), the top and bottom halves of different faces were fused to form a new facial configuration when aligned with each other. In typically developing (TD) individuals, recognition of the upper face half took more time when aligned with the bottom half than when not aligned (Carey & Diamond, 1994; Nishimura et al., 2008; Teunisse & de Gelder, 2003; Young et al., 1987). This composite effect only occurred in the upright orientation. The absence of the composite effect in adolescents with ASD indicates that they make less use of the configurational information of a face.

Moreover, Deruelle et al. (2004) found that children with ASD exhibited better performance compared with a control group when using high rather than low spatial frequency (LSF) in an identity-matching task. Previous studies (Shulman & Wilson, 1987) have demonstrated that LSF images conveyed more configural features than local ones, whereas the local features are primarily conveyed by high spatial frequency (HSF). When applied to face recognition, the low-pass filter makes facial features vague, whereas when faces are high-pass filtered, facial features appear to be emphasized. Deruelle et al. (2004) explained that the fine details of facial features (i.e., local cues) are available when the stimulus contains HSFs, but not when it contains only LSFs. Thus, the results indicated that children with ASD relied more on local (HSF) cues than on configural (LSF) cues when processing faces. This local advantage in face processing in children with ASD was discussed in line with the WCC theory (for an updated review, see Happé & Frith, 2006) and the EPF model (for an updated review, see Mottron et al., 2006). Deruelle et al. (2004) suggested that the local bias observed in children with ASD could appear in the early stages of information processing, such as spatial frequency decoding. Superior processing of face parts may explain a preference for HSFs in matching faces. Individuals with ASD might become more dependent on local cues and thereby be disadvantaged when local cues are reduced (Lahaie et al., 2006).

Facial expression processing in children with ASD

Facial expressions are important for nonverbal communication, and even neonates discriminate facial expression (Field, Woodson, Greenberg, & Cohen, 1982; Meltzoff & Moore, 1977). Some studies have revealed that children with ASD performed less accurately than controls in facial expression-matching tasks (Celani et al., 1999; Deruelle et al., 2004). Adults with ASD selectively showed a deficit in rapid emotion processing (Clark, Winkielman, & McIntosh, 2008). Rump, Giovannelli, Minshew, and Strauss (2009) found that children (aged 5–7 years) and adults with ASD performed less accurately than the control group when they were asked to identify the emotion of actors whose facial expression varied subtly in brief video clip presentations. These results suggest that the emotion recognition skills of children with ASD reach the level of those of their TD peers between the ages of 8 and 12 years and remain relatively comparable through adolescence, whereas TD children continue to develop proficiency for recognizing facial expressions through to adulthood; therefore, the control adults performed significantly better than the adults with ASD. In fact, others have reported that children (Buitelaar, van der Wees, Swaab-Barneveld, & van der Gaag, 1999; Gepner et al., 2001; Robel et al., 2004) and adolescents (Tracy, Robins, Schriber, & Solomon, 2011) with ASD do not perform significantly worse than their controls. However, facial expression processing in individuals with ASD was also somewhat different from that in TD individuals. In the sorting task, for example, children with ASD tended to sort faces by the hats and neglected the facial expressions, while TD children showed a preference for sorting by the facial expressions (Weeks & Hobson, 1987). Grossman, Klin, Carter, and Volkmar (2000) found that children with Asperger syndrome may utilize compensatory strategies such as verbal mediation to process facial expressions. To summarize, school-age children with ASD show apparently typical performance in facial expression processing in some tasks, but could be using an atypical strategy to process facial expression. Thus, it is critical to study the mechanism underlying facial expression processing in children in this age range.

Although analyses of expression and identity proceed independently (Tranel, Damasio, & Damasio, 1988; Young, McWeeny, Hay, & Ellis, 1986), TD individuals also process facial expression configurally. For example, the inversion effect (McKelvie, 1995; Valentine & Bruce, 1988) and the composite effect (Calder, 2000) have also been found in facial expression processing. In contrast, adolescents with ASD also seem to process them locally because they performed better than controls at processing images in an inverted orientation (Hobson et al., 1988). To our best knowledge, no studies have reported the composite effect of facial expressions in individuals with ASD.

Deruelle and Fagot (2005) found that TD children (5–6 years, 7–8 years) relied preferentially on HSFs in an emotion categorization task. They were asked to indicate whether the person was smiling or grimacing as a hybrid high-pass/low-pass face was presented. This HSF bias was not significant for the 10-year-old TD children (Deruelle, Rondan, Salle-Collemiche, Bastard-Rosset, & Da Fonséca, 2008) and adults (Deruelle & Fagot, 2005). In contrast, children with ASD (mean age 10 years and 5 months) showed a high-pass bias compared with controls (Deruelle et al., 2008), whereas adults with ASD showed a low-pass bias during emotion categorization (Santos, Rondan, Rosset, Da Fonséca, & Deruelle, 2008). In addition, adults with ASD were less accurate than controls in recognizing facial expressions from LSFs (Kätsyri, Saalasti, Tiippana, von Wendt, & Sams, 2008).

As reviewed above, studies on facial expression processing in ASD are nevertheless scarce compared with those examining the processing of face identity, and the results are often inconsistent between different studies. In the current study, we used spatial filtering and an inverted presentation in order to investigate whether facial expression processing in children with ASD is based primarily on local information. The combination of these two manipulations may shed new light on the nature of facial expression processing in children with ASD. For example, if children with ASD showed the inversion effect in the LSF condition, it suggests that they retain configural processing for facial expressions even though they show impairment in processing LSF images. This is the first time spatial filtering of images of facial expressions in an inverted orientation for children with ASD has been presented. There are many studies using inverted faces in individuals with ASD, but the results have not been consistent (Faja et al., 2009; Hobson et al., 1988; Lahaie et al., 2006; Langdell, 1978; Nishimura et al., 2008; Rose et al., 2007; Teunisse & de Gelder, 2003). This suggests that further replication of the inversion effect in children with ASD is needed. According to previous studies (Celani et al., 1999; Deruelle et al., 2004), children with ASD show lower performance than TD children in the broad (normal) spatial frequency (BSF) condition. We predicted that the performance of children with ASD would be less accurate than that of TD children, in the condition of upright LSF facial expression, because the LSF conveys few local cues of faces and plays a dominant role in the configural processing of faces. Furthermore, we predicted that the performance of children with ASD would not differ from the performance of TD children in the condition of upright HSF facial expression. This is because the HSF conveys many local cues, therefore making it easier to process facial features locally.

Consistent with a previous study of adolescents with ASD (Hobson et al., 1988), we also predicted that no inversion effect would be found in the BSF and the HSF conditions. If no inversion effect was found in all spatial frequencies, particularly in the LSF condition, this would indicate that children with ASD process upright facial expressions locally. In contrast, if children with ASD showed the inversion effect in the LSF condition, this would indicate that they can process facial expressions configurally when little featural information is available. For TD children we predicted that they would process facial expressions configurally, and that the inversion effect would be observed in all spatial frequencies. Note that HSF faces are not processed purely locally, because the HSF enables the extraction of fine metric relations among feature cues, that is, configural information (Goffaux, Hault, Michel, Vuong, & Rossion, 2005).

Method

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

Participants

The participants consisted of 18 children with ASD (17 male and one female; 8.4–14.5 years old; mean age = 11.9 years, SD = 1.8 years) and 19 TD children (13 male and six female; 6.4–15.3 years old; mean age = 11.4 years, SD = 2.3 years). All the children were students of a primary and junior high school for children both with and without ASD. The children with ASD had been diagnosed by at least one child psychiatrist when they enrolled in the school. In addition, after parental interviews and clinical observations, experienced clinical psychologists (Y. T. and K. Y.) confirmed the diagnoses according to DSM-IV criteria (APA, 1994). The Japanese version of Raven's Coloured Progressive Matrices (RCPM; Raven, 1956; Sugishita & Yamazaki, 1993) was administered to all the children as a measure of visuospatial intelligence (children with ASD: mean score = 30.0, SD = 5.4; TD children: mean score = 32.3, SD = 4.2). Written informed consent was obtained from the children and their parents. The study was approved by the Research Ethics Committee of the University of Tokyo.

Stimuli

The stimuli were grayscale 256 × 256-pixel photographs of adult faces (four male and four female) showing four different facial expressions (happy, angry, sad, surprised) (Kudo & Matsumoto, 1996). These eight photographs had a BSF. Following Deruelle et al. (2004), each photograph was low-pass filtered (LSF, below 2 cycles/degree of the visual angle) and high-pass filtered (HSF, above 6 cycles/degree of the visual angle) using ImageJ software, and the stimuli set contained 16 filtered faces (Figure 1). They were of 6 deg × 6 deg of the visual angle and were presented in a two-alternative forced-choice (2AFC) matching task. The distance between the target image and the probes was approximately 4.5 deg of the visual angle, and the distance between the probes was approximately 9.5 deg of the visual angle.

figure

Figure 1. Examples of the broad spatial frequency (BSF) face (top left), the low spatial frequency (LSF) face (top middle) and the high spatial frequency (HSF) face (top right). An example of the two-alternative forced-choice (2AFC) matching task (bottom); the centre image is a target and the left and right images are probes.

Download figure to PowerPoint

Procedure

The participants were tested individually in a quiet room at the University of Tokyo. The experiment was run on a laptop computer with a 14.1-in. TFT monitor. The participants were seated approximately 60 cm from the monitor to complete the 2AFC matching task. A target face was presented on one side of the monitor and a cursor was displayed on the opposite side. The participants controlled the cursor using a USB joy pad. When they moved the cursor to the target face, two probes (unfiltered BSF faces) appeared (Figure 1). The participants were asked to indicate which of the two faces showed the same facial expression as the target face by moving the cursor.

Design

The study consisted of two upright face blocks and two inverted face blocks, comprising 18 test trials for each block. Ten children with ASD and eight TD children observed the upright face block first, and the others observed the inverted face block first. Within each block, the trials were presented in a random order. In the upright face condition, there were 36 trials, which included 12 (four expressions × three distracter expressions) target faces of three types of spatial frequencies (BSF, LSF, or HSF). In half of the trials, the sex of the target image and the correct probe was the same, (i.e., the target image and the correct probe image were exactly the same because we obtained only one facial expression photograph per sex) and in the other half of the trials, the sex was different. In the inverted face condition, the stimuli of the upright face condition were shown upside-down. The experimental design consisted of one between-participants factor of group (children with ASD or TD children) and two within-participants factors of orientation (upright or inverted) and spatial frequency (BSF, LSF, or HSF).

Results

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

There was no significant group difference in the RCPM scores (t = −1.43, p > .10), indicating that the visuospatial intelligence of children with ASD and TD children did not differ. There was also no significant group difference in their chronological age (t = .71, p > .40). We analyzed the correct rate on the upright and inverted face tasks.

The upright face task

Figure 2 and Table 1 show the correct rate on the upright face tasks by the two groups for the three conditions. The distribution of the correct rate made the use of parametric tests inappropriate. Therefore, nonparametric tests were used to examine these data. Bonferroni corrections were adopted for all multiple comparisons. The children with ASD exhibited lower performance than the TD children in the BSF and LSF condition (BSF: Z = −3.05, p < .01; LSF: Z = −2.22, p < .05, Mann-Whitney U-test). In the HSF condition, the group difference was not significant (Z = −1.33, p > .30).

figure

Figure 2. Mean correct rates for three spatial frequency conditions and two orientations in children with ASD and TD children. Error bar = standard error. ASD = autism spectrum disorder; BSF = broad spatial frequency; HSF = high spatial frequency; LSF = low spatial frequency; TD = typically developing. *p < .05. **p < .01.

Download figure to PowerPoint

Table 1. Mean correct rates and standard deviations (M ± SD) of each condition (%)
 Broad spatial frequencyLow spatial frequencyHigh spatial frequency
Children with autism spectrum disorder (n = 18)
Upright94.0 ± 7.4683.8 ± 11.9593.5 ± 10.90
Inverted88.4 ± 14.6168.5 ± 16.5689.8 ± 11.27
Typically developing children (n = 19)
Upright99.6 ± 1.9192.1 ± 8.0997.8 ± 5.44
Inverted95.6 ± 6.4475 ± 17.7992.5 ± 9.97

The condition effect was significant in both groups (children with ASD: χ2 = 11.1, p < .01; TD children: χ2 = 15.1, p < .01, Friedman test), and a Wilcoxon test with Bonferroni correction was performed in each group. Within children with ASD, the performance in the LSF condition was significantly lower than in the BSF and the HSF condition (LSF-BSF: Z = −2.84, p < .05; LSF-HSF: Z = −2.74, p < .05). There was no significant difference between the BSF and the HSF condition (Z = –.16, p > .80). Within TD children, performance in the LSF condition was significantly lower than in the BSF condition (Z = −3.02, p < .05). There was a marginal difference between the LSF and the HSF condition (Z = −2.20, p = .08). There was no significant difference between the BSF and HSF conditions (Z = −1.30, p > .50).

The inverted face task

Figure 2 and Table 1 show the correct rate on the inverted face tasks by the two groups for the three conditions. A marginally significant difference was found between children with ASD and TD children in the BSF condition (Z = −1.76, p = .08, Mann-Whitney U-test). In the LSF and HSF condition, the group difference was not significant (LSF: Z = −1.08, p > .20; HSF: Z = –.83, p > .40).

The condition effect was significant in children with ASD (χ2 = 23.3, p < .01, Friedman test) and in TD children (χ2 = 19.9, p < .01), and a Wilcoxon test with Bonferroni correction was performed in each group. Within children with ASD, performance in the LSF condition was significantly lower than in the BSF and HSF condition (LSF-BSF: Z = −3.29, p < .01; LSF-HSF: Z = −3.44, p < .01). There was no significant difference between the BSF and HSF condition (Z = –.24, p > .80). Within TD children, performance in the LSF condition was significantly lower than in the BSF and the HSF condition (LSF-BSF: Z = −3.43, p < .01; LSF-HSF: Z = −2.70, p < .05). There was no significant difference between the BSF and the HSF conditions (Z = –.94, p > .30).

The inversion effect

In each group, a Wilcoxon test was performed. Within TD children, performance with inverted faces was significantly lower than with upright faces in each spatial frequency condition (BSF: Z = −2.53, p < .05; LSF: Z = −3.23, p < .01; HSF: Z = −2.39, p < .05). Within children with ASD, the difference between upright faces and inverted faces in the BSF condition did not reach significance (Z = −1.90, p = .06). In the LSF condition, performance with inverted faces was significantly lower than with upright faces (Z = −2.90, p < .01). In the HSF condition, there was no significant difference between upright faces and inverted faces (Z = –.99, p > .30).

Correlation

We calculated the average correct rate of all conditions in each participant. Figure 3a shows that in general, children with ASD performed better with age (ρ = .59, p < .05), while TD children did not (ρ = .34, p > .10). In children with ASD, there was a significant correlation between performance in the LSF condition and age with both the upright (ρ = .58, p < .05; Figure 3b) and inverted (ρ = .61, p < .01) faces. In addition, in TD children, the correlation of performance in the LSF condition and age was marginally significant with upright faces (ρ = .44, p = .06; Figure 3b). No significant correlations were found between other conditions and age (all ρ < .46, p > .05).

figure

Figure 3. (a) Scatterplot of the age and the average correct rate of all conditions in each participant. (b) Scatterplot of the age and the correct rate of the upright LSF condition in each participant. ASD = autism spectrum disorder; LSF = low spatial frequency; TD = typically developing.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

First, as predicted, children with ASD showed lower performance than TD children in the LSF condition with upright faces. Because the LSF images convey more configural rather than local information, this result suggests that children with ASD had more difficulty in processing facial expression configurally than did TD children. This finding is consistent with the research of Kätsyri et al. (2008), in which adults with ASD were less accurate than controls in recognizing facial expressions from LSFs. Children with ASD also showed lower performance than TD children in the BSF condition, and this is consistent with some previous studies using a matching-to-sample paradigm (Celani et al., 1999; Deruelle et al., 2004). In contrast, the two groups showed no difference in the HSF condition. In the HSF images, in which local cues are more easily accessible due to the relative absence of LSF information, the children with ASD processed facial expression as well as did TD children. As the group difference was found not in the HSF but in the BSF condition, the latter of which includes the same amount of HSF components as the former, children with ASD might have difficulties in extracting local cues from the BSF image, which also contains rich LSF components compared to the HSF image. This would cause children with ASD to be more dependent on local cues. These results indicate an HSF preference in children with ASD, which is consistent with previous studies (Deruelle et al., 2004, 2008).

Second, while TD children showed the inversion effect in both the BSF and the HSF condition, children with ASD did not. The results of the TD children replicated previous findings in which facial expression processing became difficult when faces were presented in an inverted compared with an upright orientation (McKelvie, 1995; Valentine & Bruce, 1988). It should be noted that HSF faces are not processed purely locally, as the configural information is also available, to some extent, from the HSF faces. Because TD children processed the HSF faces configurally, the inversion effect in the HSF condition was also observed. Meanwhile, the absence of the inversion effect for the BSF and HSF conditions in children with ASD suggests local rather than configural processing for facial expressions because only local information is available in inverted faces (Carey & Diamond, 1994). Moreover, in the inverted condition, there was no significant difference between the children with ASD and TD children in any spatial frequency condition. This result contrasts with that of the upright condition, in which a group difference was found in both the BSF and the LSF conditions. The current result of the noninversion effect for the BSF and HSF faces in children with ASD is consistent with the previous study of adolescents with ASD (Hobson et al., 1988).

In contrast, in the LSF condition, children with ASD showed the inversion effect just like TD children, which is the novel finding of the current study. If children with ASD had no ability to process configural information, they should not show the inversion effect even in the LSF condition. Even if they process facial expressions from the subtle local cues remaining in the LSF images, the inversion effect should not be observed at all, because the inversion effect does occur only when the upright facial expressions are processed configurally. Therefore, these results indicate that configural processing of facial expressions may not be completely absent in children with ASD, but that they tend to process it locally when images have rich local information. These results are consistent with the recent WCC theory (Happé, 1999; Happé & Frith, 2006) and the EPF model (Mottron et al., 2006). The current study demonstrates the capacity for configural processing of facial expressions in children with ASD.

It was more difficult for both TD children and children with ASD to respond to the LSF faces compared with the BSF and HSF faces. This is consistent with the results of Leonard, Karmiloff-Smith, and Johnson (2010) on the performance of TD 9- and 10-year-olds. The performance of TD adults was also significantly worse when the middle spatial frequencies of an upright face were masked than when only the LSFs or HSFs were masked. In addition, the recognition of the LSF face images was less accurate than that of the HSF images in TD adults (Fiorentini, Maffei, & Sandini, 1983; Nagayama, Yoshida, & Toshima, 1995).

In children with ASD who have difficulty in processing LSF images, which retain configural and little local information, there was a correlation between their age and their performance in the LSF condition. This tendency was also found in TD children. These results might suggest that the skill of configural processing for faces develops with age. Children with ASD (Leonard, Annaz, Karmiloff-Smith, & Johnson, 2011) and TD children, at least those under 10 years old (Leonard et al., 2010, 2011) tended to rely on the HSFs for face recognition. In addition, children aged less than approximately 10 years did not process faces configurally (Carey & Diamond, 1977), and adult expertise in configural processing is slow to develop (Mondloch, Le Grand, & Maurer, 2002). The correlation between performance in the LSF condition and age was consistent with these previous studies.

In this study, we could not test which part of the face children with ASD used as a cue to process the facial expressions. On identity tasks, some studies have reported that the mouth was a cue for children with ASD, while it was the eyes for TD children (Joseph & Tanaka, 2003; Langdell, 1978). Their ability to identify faces on the basis of mouth cues but not eye cues could extend to the recognition of facial expressions (Joseph & Tanaka, 2003). Contrary to these arguments, Rutherford and Towns (2008) used an eye-tracking device to demonstrate that adults with ASD, like TD adults, looked longer at the eyes than the mouth when identifying facial expressions. Further investigation of this issue is required, possibly with a composite paradigm (Calder, 2000).

What might account for the part-based face processing in children with ASD? Some argue that the developmental origin of atypical face processing in individuals with ASD is the result of insufficient facial orienting, which is possibly due to the lack of attentional bias (Schultz, 2005) or motivation (Dawson, Webb, & McPartland, 2005) toward others' faces. In fact, infants and children with ASD lack an attentional bias toward others' faces (Kikuchi, Senju, Tojo, Osanai, & Hasegawa, 2009; Osterling, Dawson, & Munson, 2002). Future studies should investigate whether an increased amount of attention to faces could lead to face expertise and possibly trigger holistic/configural face processing in children with ASD during their development.

Several limitations of the current study should be noted. First, we could not investigate the relation between each facial expression and the spatial frequency because of the small number of trials used in this study. In addition, as is often the case with studies that involve atypically developing children, the interpretation of the null results has to be treated carefully due to the smaller number of trials compared with adult studies, which is essential to minimize task loads but which could increase the within-participant variability of the data. Further studies, ideally with adult participants, who can tolerate longer experimental sessions than children, would be beneficial to examine these questions. Second, as we used a joypad for the children's responses, we could not analyze the reaction time with sufficient temporal resolution. Further studies will be required to test whether the inversion effect, which was observed in the accuracy data, could also be observed with reaction time data. Third, the limited age range of the children with ASD (8–14 years old) does not allow for analyses of the developmental trajectory in earlier and later age ranges. As we found a positive correlation between performance and age in children with ASD, further studies would be fruitful to follow up development in a wider age range. Fourth, we followed the technique used by Deruelle et al. (2004) and Costen, Parker, and Craw (1996) for the application of spatial filtering, although the units were different across studies. Some researchers suggest that it might be relatively easy to identify faces from the LSF images in the latter study (Endo & Kirita, 2004), and this might affect the results. We agree that researchers would benefit from reaching a consensus on the useful methodology for future investigations to promote ease of comparison across studies (Endo & Kirita, 2004; Leonard et al., 2010).

In summary, we investigated facial expression processing in children with ASD by using LSF (facilitating configural processing), HSF (facilitating local processing), and BSF (normal) images. In a two-alternative forced-choice matching task, children with ASD performed less accurately in the BSF and LSF conditions, but not in the HSF condition with the upright orientation. In addition, both TD children and children with ASD showed the inversion effect in the LSF condition, whereas the inversion effects in the BSF and the HSF condition did not reach significance in children with ASD, unlike TD children. These results suggest a local processing preference for facial expressions in children with ASD.

References

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References
  • American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association.
  • Boucher, J., Lewis, V., & Collis, G. (1998). Familiar face and voice matching and recognition in children with autism. Journal of Child Psychology and Psychiatry, 39, 171181.
  • Buitelaar, J. K., van der Wees, M., Swaab-Barneveld, H., & van der Gaag, R. (1999). Verbal memory and performance IQ predict theory of mind and emotion recognition ability in children with autistic spectrum disorders and psychiatric control children. Journal of Child Psychology and Psychiatry, 40, 869881.
  • Calder, A. J. (2000). Configural information in facial expression perception. Journal of Experimental Psychology, 26, 527551.
  • Carey, S., & Diamond, R. (1977). From piecemeal to configurational representation of faces. Science, 195, 213313.
  • Carey, S., & Diamond, R. (1994). Are faces perceived as configurations more by adults than by children? Visual Cognition, 1, 253274.
  • Celani, G., Battacchi, M. W., & Arcidiacono, L. (1999). The understanding of the emotional meaning of facial expressions in people with autism. Journal of Autism and Developmental Disorders, 29, 5766.
  • Clark, T. F., Winkielman, P., & McIntosh, D. N. (2008). Autism and the extraction of emotion from briefly presented facial expressions: Stumbling at the first step of empathy. Emotion, 8, 803809.
  • Costen, N. P., Parker, D. M., & Craw, I. (1996). Effects of high-pass and low-pass spatial filtering on face identification. Perception and Psychophysics, 58, 602612.
  • Davies, S., Bishop, D., Manstead, A. S. R., & Tantam, D. (1994). Face perception in children with autism and Asperger's syndrome. Journal of Child Psychology and Psychiatry, 35, 10331057.
  • Dawson, G., Webb, S. J., & McPartland, J. (2005). Understanding the nature of face processing impairment in autism: Insights from behavioral and electrophysiological studies. Developmental Neuropsychology, 27, 403424.
  • Deruelle, C., & Fagot, J. (2005). Categorizing facial identities, emotions, and genders: Attention to high- and low-spatial frequencies by children and adults. Journal of Experimental Child Psychology, 90, 172184.
  • Deruelle, C., Rondan, C., Gepner, B., & Tardif, C. (2004). Spatial frequency and face processing in children with autism and Asperger syndrome. Journal of Autism and Developmental Disorders, 34, 199210.
  • Deruelle, C., Rondan, C., Salle-Collemiche, X., Bastard-Rosset, D., & Da Fonséca, D. (2008). Attention to low- and high-spatial frequencies in categorizing facial identities, emotions and gender in children with autism. Brain and Cognition, 66, 115123.
  • Endo, M., & Kirita, T. (2004). Role of spatial frequency information in face recognition. Human Science: Bulletin of the Faculty of Law and Letters, University of the Ryukyus, 13, 179202. (In Japanese with English abstract.)
  • Faja, S., Webb, S. J., Merkle, K., Aylward, E., & Dawson, G. (2009). Brief report: Face configuration accuracy and processing speed among adults with high-functioning autism spectrum disorders. Journal of Autism and Developmental Disorders, 39, 532538.
  • Field, T. M., Woodson, R., Greenberg, R., & Cohen, D. (1982). Discrimination and imitation of facial expression by neonates. Science, 218, 179181.
  • Fiorentini, A., Maffei, L., & Sandini, G. (1983). The role of high spatial frequencies in face perception. Perception, 12, 195201.
  • Gepner, B., Deruelle, C., & Grynfeltt, S. (2001). Motion and emotion: A novel approach to the study of face processing by young autistic children. Journal of Autism and Developmental Disorders, 31, 3745.
  • Gliga, T., & Csibra, G. (2007). Seeing the face through the eyes: A developmental perspective on face expertise. Progress in Brain Research, 164, 323339.
  • Goffaux, V., Hault, B., Michel, C., Vuong, Q. C., & Rossion, B. (2005). The respective role of low and high spatial frequencies in supporting configural and featural processing of face. Perception, 34, 7786.
  • Grossman, J. B., Klin, A., Carter, A. S., & Volkmar, F. R. (2000). Verbal bias in recognition of facial emotions in children with Asperger syndrome. Journal of Child Psychology and Psychiatry, 41, 369379.
  • Happé, F. (1999). Autism: Cognitive deficit or cognitive style? Trends in Cognitive Sciences, 3, 216222.
  • Happé, F., & Frith, U. (2006). The weak coherence account: Detail-focused cognitive style in autism spectrum disorders. Journal of Autism and Developmental Disorders, 36, 525.
  • Hobson, R. P., Ouston, J., & Lee, A. (1988). What's in a face? The case of autism. British Journal of Psychology, 79, 441453.
  • Joseph, R. T., & Tanaka, J. (2003). Holistic and part-based face recognition in children with autism. Journal of Child Psychology and Psychiatry, 44, 529542.
  • Kätsyri, J., Saalasti, S., Tiippana, K., von Wendt, L., & Sams, M. (2008). Impaired recognition of facial emotions from low-spatial frequencies in Asperger syndrome. Neuropsychologia, 46, 18881897.
  • Kikuchi, Y., Senju, A., Tojo, Y., Osanai, H., & Hasegawa, T. (2009). Faces do not capture special attention in children with autism spectrum disorder: A change blindness study. Child Development, 80, 14211433.
  • Klin, A., Sparrow, S. S., de Bildt, A., Cicchetti, D. V., Cohe, D. J., & Volkmar, F. R. (1999). A normed study of face recognition in autism and related disorders. Journal of Autism and Developmental Disorders, 29, 499508.
  • Kudo, T., & Matsumoto, D. (1996). Nihonjin no kanjo sekai [Unmasking Japan: Myths and realities about the emotions of the Japanese]. Tokyo: Seishin Shobo. (In Japanese.)
  • Lahaie, A., Mottron, L., Arguin, M., Berthiaume, C., Jemel, B., & Saumier, D. (2006). Face perception in high-functioning autistic adults: Evidence for superior processing of face parts, not for a configural face-processing deficit. Neuropsychology, 20, 3041.
  • Langdell, T. (1978). Recognition of faces: An approach to the study of autism. Journal of Child Psychology and Psychiatry, 19, 255268.
  • Leonard, H. C., Annaz, D., Karmiloff-Smith, A., & Johnson, M. H. (2011). Brief report: Developing spatial frequency biases for face recognition in autism and Williams syndrome. Journal of Autism and Developmental Disorders, 41, 968973.
  • Leonard, H. C., Karmiloff-Smith, A., & Johnson, M. H. (2010). The development of spatial frequency biases in face recognition. Journal of Experimental Child Psychology, 106, 193207.
  • McKelvie, S. J. (1995). Emotional expression in upside-down faces: Evidence for configurational and componential processing. British Journal of Social Psychology, 34, 325334.
  • Meltzoff, A. N., & Moore, M. K. (1977). Imitation of facial and manual gestures by human neonates. Science, 198, 7578.
  • Mondloch, C., Le Grand, R., & Maurer, D. (2002). Configural face processing develops more slowly than featural face processing. Perception, 31, 553566.
  • Mottron, L., Burack, J. A., Iarocci, G., Belleville, S., & Enns, J. T. (2003). Locally oriented perception with intact global processing among adolescents with high-functioning autism: Evidence from multiple paradigms. Journal of Child Psychology and Psychiatry, 44, 904913.
  • Mottron, L., Dawson, M., Soulières, I., Hubert, B., & Burack, J. (2006). Enhanced perceptual functioning in autism: An update, and eight principles of autistic perception. Journal of Autism and Developmental Disorders, 36, 2743.
  • Nagayama, R., Yoshida, H., & Toshima, T. (1995). Interrelationship between the facial expression and familiarity: Analysis using spatial filtering and inverted presentation. Japanese Journal of Psychology, 66, 327335. (In Japanese with English abstract.)
  • Nishimura, M., Rutherford, M. D., & Maurer, D. (2008). Converging evidence of configural processing of faces in high-functioning adults with autism spectrum disorders. Visual Cognition, 16, 859891.
  • Osterling, J. A., Dawson, G., & Munson, J. A. (2002). Early recognition of 1-year-old infants with autism spectrum disorder versus mental retardation. Development and Psychopathology, 14, 239251.
  • Pellicano, E., Gibson, L., Maybery, M., Durkin, K., & Badcock, D. R. (2005). Abnormal global processing along the dorsal visual pathway in autism: A possible mechanism for weak visuospatial coherence? Neuropsychologia, 43, 10441053.
  • Plaisted, K., Swettenham, J., & Rees, L. (1999). Children with autism show local precedence in a divided attention task and global precedence in a selective attention task. Journal of Child Psychology and Psychiatry, 40, 733742.
  • Raven, J. C. (1956). Coloured progressive matrices. London: Lewis.
  • Robel, L., Ennouri, K., Piana, H., Vaivre-Douret, L., Perier, A., Flament, M. F., & Mouren-Simeoni, M. C. (2004). Discrimination of face identities and expressions in children with autism: Same or different? European Child and Adolescent Psychiatry, 13, 227233.
  • Rose, F. E., Lincoln, A. J., Lai, Z., Ene, M., Searcy, Y. M., & Bellugi, U. (2007). Orientation and affective expression effects on face recognition in Williams syndrome and autism. Journal of Autism and Developmental Disorders, 37, 513522.
  • Rump, K. M., Giovannelli, J. L., Minshew, N. J., & Strauss, M. S. (2009). The development of emotion recognition in individuals with autism. Child Development, 80, 14341447.
  • Rutherford, M. D., & Towns, A. M. (2008). Scan path differences and similarities during emotion perception in those with and without autism spectrum disorders. Journal of Autism and Developmental Disorders, 38, 13711381.
  • Santos, A., Rondan, C., Rosset, D. B., Da Fonséca, D. D., & Deruelle, C. (2008). Mr. Grimace or Ms. Smile. Does categorization affect perceptual processing in autism? Psychological Science, 19, 7076.
    Direct Link:
  • Sasson, N. J. (2006). The development of face processing in autism. Journal of Autism and Developmental Disorders, 36, 381394.
  • Schultz, R. T. (2005). Developmental deficits in social perception in autism: The role of the amygdala and fusiform face area. International Journal of Developmental Neuroscience, 23, 125141.
  • Shah, A., & Frith, U. (1983). An islet of ability in autistic children: A research note. Journal of Child Psychology and Psychiatry, 24, 613620.
  • Shulman, G. L., & Wilson, J. (1987). Spatial frequency and selective attention to local and global information. Perception, 16, 89101.
  • Sugishita, M., & Yamazaki, Y. (1993). Raven shikisai matorikkusu kensa [Japanese Raven's coloured progressive matrices]. Tokyo: Nihon Bunka Kagakusha. (In Japanese.)
  • Teunisse, J., & de Gelder, B. (2003). Face processing in adolescents with autistic disorder: The inversion and composite effects. Brain and Cognition, 52, 285294.
  • Tracy, J. L., Robins, R. W., Schriber, R. A., & Solomon, M. (2011). Is emotion recognition impaired in individuals with autism spectrum disorders? Journal of Autism and Developmental Disorders, 41, 102109.
  • Tranel, D., Damasio, A. R., & Damasio, H. (1988). Intact recognition of facial expression, gender, and age in patients with impaired recognition of face identity. Neurology, 38, 690696.
  • Valentine, T., & Bruce, V. (1988). Mental rotation of faces. Memory and Cognition, 16, 556566.
  • Volkmar, F. R., Sparrow, S. S., Rende, R. D., & Cohen, D. J. (1989). Facial perception in autism. Journal of Child Psychology and Psychiatry, 30, 591598.
  • Weeks, S. J., & Hobson, R. P. (1987). The salience of facial expression for autistic children. Journal of Child Psychology and Psychiatry, 28, 137152.
  • White, S. J., & Saldaña, D. (2011). Performance of children with autism on the embedded figures test: A closer look at a popular task. Journal of Autism and Developmental Disorders, 41, 15651572.
  • Yin, R. K. (1969). Looking at upside-down faces. Journal of Experimental Psychology, 81, 141145.
  • Young, A. W., Hellawell, D., & Hay, D. C. (1987). Configurational information in face perception. Perception, 16, 747759.
  • Young, A. W., McWeeny, K. H., Hay, D. C., & Ellis, A. W. (1986). Matching familiar and unfamiliar faces on identity and expression. Psychological Research, 48, 6368.