تطبیق مدال داخلی و متقابل معین ابراز هیجانی در کودکان با اختلالات طیف اوتیسم

Abstract Children with autism spectrum disorders (ASDs) exhibit difficulties in their comprehension of others’ emotions. According to a variety of experimental procedures, many of them can be classified into two types according to the modality of their stimuli and responses. These are intramodal (visual stimulus–visual stimulus) and cross-modal (auditory stimulus–visual stimulus) matching. Previous studies tested both intramodal and cross-modal matching only in adolescents with ASD, although young children with ASD have also been found to have difficulties with cross-modal matching but not intramodal matching. The purpose of this study was to compare the intramodal and cross-modal matching of emotional expression in young children with ASD and typically developing (TD) children. Ten children with ASD aged 4–8 and 22 developmental age (DA)-matched TD children aged 3–6 participated in this study. Pictures of facial expressions were used as visual stimuli, and affective prosodies were used as auditory stimuli. The results showed that the children with ASD were less accurate than the TD children in cross-modal matching but equally accurate on intramodal matching. These findings are discussed along with the modality of stimuli and responses, and the ages of the participants.

1. Introduction Autism spectrum disorders (ASDs) are neurodevelopmental disorders characterized by distinct social deficits (American Psychiatric Association, 2013). Individuals with ASD have difficulties particularly in their comprehension of others’ emotions conveyed from nonverbal cues, such as facial expressions or affective prosodies (e.g., Braverman et al., 1989, Hobson, 1986a, Hobson, 1986b, Macdonald et al., 1989 and Tantam et al., 1989). However, the results of previous studies have yielded inconsistent findings. Some studies found that individuals with ASD face difficulties in understanding other's emotions (e.g., Bormann-Kischkel et al., 1995, Celani et al., 1999 and Davies et al., 1994). Other studies did not find such deficits (e.g., Capps et al., 1992, Gepner et al., 2001 and Prior et al., 1990). Matsuda and Yamamoto, 2013 and Matsuda and Yamamoto, 2014 suggested that these inconsistencies occurred due to variations in the experimental tasks and ages of the participants. Previous emotion comprehension studies can be classified into two types according to the modality of their stimuli and responses. These are intramodal (i.e., visual stimulus–visual stimulus) and cross-modal (i.e., auditory stimulus–visual stimulus) matching. In some studies where intramodal matching was used, pictures or movies showing facial expressions were presented as sample stimuli, and either facial expressions or texts were used as response choices. In other studies where cross-modal matching was used, affective prosodies representing different emotional states were presented as sample stimuli, and illustrations, texts, or both were used as response choices. To date, few studies have used both intramodal and cross-modal matching (Jones et al., 2011 and Loveland et al., 1997). Jones et al. (2011) tested adolescents with ASD (mean age: 15 years, 6 months), and as did Loveland et al. (1997) (mean age: 12 years, 2 months). In both studies, the results showed no difference in intramodal and cross-modal matching between adolescents with ASD and typically developing (TD) adolescents. However, to our knowledge, no study has compared the intramodal and cross-modal matching of emotional expressions in younger children with ASD. It is possible that the difficulties children with ASD experience in cross-modal matching of emotional expressions are related to their age. Studies of older children with ASD, aged 12–16 years, have found that they have no difficulty in cross-modal matching (Baker et al., 2010, Grossman et al., 2010, Jones et al., 2011 and Paul et al., 2005). Meanwhile, studies of younger children with ASD, aged 9–10 years, have reported significant group differences (Boucher et al., 2000 and Peppé et al., 2007). These results suggest that younger children with ASD might have difficulties in cross-modal matching of emotional expressions. On the other hand, in intramodal matching of emotional expressions, children with ASD might have no difficulties regardless of their age. Previous studies on intramodal matching have reported that both younger (aged 6–11 years; Fink et al., 2014, Grossman et al., 2000, Lacroix et al., 2009 and Rosset et al., 2008) and older (aged 12–15 years; Castelli, 2005 and Loveland et al., 1997) children with ASD, and TD children performed equally well. These results indicate that intramodal matching is not difficult for either young or old children with ASD. The purpose of the present study was to compare intramodal and cross-modal matching of emotional expressions in young children with ASD and TD children. Ten boys with ASD (mean age: 6 years, 2 months) participated in this study. They were younger and had lower developmental ages (DAs) than the participants in previous studies. The intramodal matching required the children to match pictures of facial expressions across different facial identities. The cross-modal matching required them to match an affective prosody to the corresponding picture of the facial expression. We used four basic emotions, happy, surprised, angry, and sad, for both intramodal and cross-modal matching.

3. Results All children met the criteria to start the test sessions, based on the training session of intramodal and cross-modal matching. 3.1. Group differences for two matching tasks Fig. 1 shows a box plot comparison of intramodal and cross-modal matching for children with ASD and TD children. Box plot display for the percentages of correct responses in each matching task. ... Fig. 1. Box plot display for the percentages of correct responses in each matching task. The lines in the box indicate the median of the distribution. The upper and lower boundaries of the boxes indicate the 25th and 75th percentiles, respectively, and the upper and lower whiskers indicate highest and smallest, respectively. Figure options 3.1.1. Intramodal matching In the intramodal matching, the mean percentage of correct responses was 85.0% (SD = 29.0) for children with ASD and 98.9% (SD = 3.1) for TD children. There was no difference between the two groups in terms of the percentage of correct responses (Mann–Whitney test: Z = 1.75, p = 0.081, r = 0.31). 3.1.2. Cross-modal matching In the cross-modal matching, the mean percentage of correct responses was 58.1% (SD = 26.8) for children with ASD and 80.1% (SD = 21.4) for TD children. There was a significant difference between the ASD and TD groups (Mann–Whitney test: Z = 2.36, p = 0.018, r = 0.42). 3.2. Differences between matching tasks in each group The percentage of correct responses in intramodal matching was higher than in cross-modal matching for both children with ASD (Wilcoxon signed-rank test: Z = −2.25, p = 0.024, r = 0.71) and TD children (Wilcoxon signed-rank test: Z = −3.18, p = 0.001, r = 0.68). 3.3. Group differences for four emotions 3.3.1. Intramodal matching In the intramodal matching, the children with ASD had the following mean percentages of correct responses: 82.5% (SD = 31.3) for “happy,” 87.5% (SD = 31.7) for “angry,” 87.5% (SD = 27.0) for “sad,” and 82.5% (SD = 31.3) for “surprised.” The percentages for the TD children were as follows: 98.9% (SD = 3.1) for “happy,” 100.0% (SD = 0.0) for “angry,” 97.7% (SD = 7.4) for “sad,” and 98.9% (SD = 5.3) for “surprised” (see Table 1). There was no significant difference between the two groups with Bonferroni correction (happy: Z = 2.05, p = 0.040, r = 0.36; angry: Z = 2.13, p = 0.033, r = 0.38; sad: Z = 0.91, p = 0.322, r = 0.18; surprised: Z = 2.05, p = 0.040, r = 0.36). Table 1. Confusion matrix for each of the four emotions in each group with responses in percentages for all sample emotions. Percentages of correct responses are underlined. Group Response Intramodal matching (facial expression–facial expression) Cross-modal matching (affective prosody–facial expression) Sample Sample Happy Surprised Angry Sad Happy Surprised Angry Sad ASD Happy 82.5 2.5 7.5 2.5 52.5 25.0 17.5 5.0 Surprised 5.0 82.5 0.0 7.5 32.5 40.0 22.5 2.5 Angry 7.5 5.0 87.5 2.5 7.5 15.0 55.0 7.5 Sad 5.0 7.5 5.0 87.5 7.5 17.5 5.0 85.0 TD Happy 98.9 0.0 0.0 1.1 75.0 12.5 9.1 1.1 Surprised 1.1 98.9 0.0 1.1 10.2 68.2 6.8 1.1 Angry 0.0 0.0 100.0 0.0 5.7 9.1 79.5 0.0 Sad 0.0 1.1 0.0 97.7 9.1 10.2 4.5 97.7 Table options 3.3.2. Cross-modal matching In the cross-modal matching, the children with ASD had the following mean percentages of correct responses: 52.5% (SD = 46.3) for “happy,” 55.0% (SD = 35.0) for “angry,” 85.0% (SD = 26.9) for “sad,” and 40.0% (SD = 37.6) for “surprised.” The percentages for the TD children were as follows: 75.0% (SD = 34.5) for “happy,” 79.5% (SD = 31.5) for “angry,” 97.7% (SD = 7.4) for “sad,” and 68.2% (SD = 37.1) for “surprised” (see Table 1). There was no significant difference between the two groups with Bonferroni correction (happy: Z = 1.27, p = 0.205, r = 0.22; angry: Z = 2.10, p = 0.036, r = 0.37; sad: Z = 1.61, p = 0.108, r = 0.28; surprised: Z = 1.81, p = 0.070, r = 0.32). 3.4. Differences among the four emotions in each group 3.4.1. Intramodal matching Wilcoxon signed-rank tests showed no difference among the four emotions in intramodal matching for both children with ASD and TD children. 3.4.2. Cross-modal matching For the children with ASD, Wilcoxon signed-rank tests showed a trend that the percentage of correct responses for sad was higher than that for the other three emotions; however, there was no significant difference with Bonferroni correction (happy vs. sad: Z = −1.79, p = 0.073, r = 0.57; surprised vs. sad: Z = −2.38, p = 0.017, r = 0.75; angry vs. sad: Z = −2.41, p = 0.016, r = 0.76). For the TD children, Wilcoxon signed-rank tests showed significant differences for happy vs. sad (Z = −2.69, p = 0.007, r = 0.57) and surprised vs. sad (Z = −3.10, p = 0.002, r = 0.66). There was also a trend toward the higher percentage of correct responses for sad than angry, but the difference did not remain significant with Bonferroni correction (Z = −2.40, p = 0.016, r = 0.51).