| Summary: | 碩士 === 中國醫藥大學 === 生物醫學研究所碩士班 === 106 === Successful recognition of faces and objects is fundamental to survival. Previous studies showed that individuals with ASD adopted a featural-based rather than configural-based processing to distinguish among faces and objects. However, a direct test of whether individuals with ASD adopt any holistic processing is lacking. The present study directly tested holistic face processing and investigated face and object discrimination with equated task difficulties in adults with Asperger syndrome (AS) and neuraltypical controls (NT). A total of 25 adults with AS and 26 NT adults matched in age, gender, and years of education were tested. Experiment 1 was the composite face task where we measured the magnitude of composite face effect by comparing the performances of the aligned, inverted, and misaligned conditions. Experiment 2 was the 2AFC morphing face/object discrimination task where we recorded the accuracy and response time at five different morph levels. Moreover, we estimated the discrimination thresholds for both groups with Weilbull function curve fitting. The participants also received the paper-and-pencil Chinese Autism-spectrum Quotient questionnaire (AQ) and Social Interaction Anxiety Scale (SIAS).
Our results showed that: (1) Adults with AS had higher scores on AQ and SIAS than those of NT adults. (2) Both groups exhibited a composite face effect but the magnitude in AS group was significantly smaller. (3) For the morphed face discrimination task, the mean accuracies were not significantly different between the two groups; however, the estimated discrimination threshold in AS group was significantly higher than the NT group. (4) For the morphed object discrimination task, both groups have the same mean accuracies and estimated discrimination thresholds. In summary, our findings suggest that individuals with AS adopt a weaker holistic processing strategy when viewing faces. Moreover, AS group exhibited a larger discrimination threshold on the morphed face, and rely more on featural-based processing strategy when distinguishing among morphed objects.
|
|---|
