Objectives: This study investigated the development of ToM in Japanese students with typical development (TD) and in those with high-functioning autism spectrum disorder (ASD) applying this ToM test.
Methods: Elementary school children from the second to fourth grades (age 8–10), including 70 TD children and 63 ASD children with normal intelligence, participated in the study. Among the ASD children, the average verbal IQ was 106.5 (range 85–138) and the average performance IQ was 102.8 (range 85–122) on the WISC-III. The ToM tasks were presented to the participants via audio, letters and animations using a PC display and speaker. The passing number of the tasks were scored (0–5). To pass each task, all sub-questions including control questions had to be correctly answered.
Results: The average scores (standard deviation) of the ToM tasks were 4.32 (0.85), 4.54 (0.66) and 4.71 (0.46) for TD children, and 2.32 (1.46), 3.29 (1.37) and 4.21 (1.12) for ASD children of the second, third and fourth grades, respectively. A two-way analysis of variance (ANOVA) (3 (Grade: second versus third versus fourth) × 2 (Group: ASD versus TD)) was conducted. The interaction between age and presence of ASD was significant (F (2,127) = 5.11, p < .01). The simple main effect of grade was significant only in ASD children (F (2,127) = 14.92, p < .001). The ToM performance in the second grade was significantly lower than that in the third grade, and was significantly lower in the third grade than that in the fourth grade. The simple main effect of the presence of ASD was significant in the second (F (1,127) = 44.58, p < .001) and third grades (F (1,127) = 44.58, p < .001). The ToM performance in ASD children was significantly lower than that in TD children in the second and third grades. The logistic regression analysis indicated that the ToM score was a significant predictor of ASD in the second (B = −1.38, Wald = 12.97, p < .001, odds ratio = 0.25) and third grades (B = −1.55, Wald = 9.67, p < .01, odds ratio = 0.21), but was not significant in the fourth grade. The predictive values of ASD were 78% and 73% in the second and third grades, respectively.
Conclusions: These results suggest that the animated version of the ToM test can detect ToM problems in ASD children until the third grade. However, it cannot accurately detect these problems in the fourth grade. Therefore, the results on this ToM test should be evaluated carefully, particularly when applied as an assessment tool among children in the fourth grade or higher. Happé (1995) suggested that the verbal mental age of nine is critical for passing ToM tasks. Our results also support this finding.
See more of: Cognition and Behavior
See more of: Symptoms, Diagnosis & Phenotype