過去研究文獻指出威廉氏症候群有見樹不見林的認知處理傾向，不僅在非語言範疇觀察到，近年來在語言範疇裡也得到證明，符合自閉症症候群中心連貫缺損認知處理模式，本研究旨在發掘威廉氏症候群臉部辨識能力是否也有中心連貫缺損情形，用誘發事件相關電位(event-related potentials, ERPs)尋找整體訊息處理歧異神經生理證據。研究方法讓受試者看連續呈現在電腦螢幕上的臉部圖片，第一張為模型臉(model face)，第二張為改變臉，有特徵改變(feature-changed)及整體改變(configure-changed)兩種臉，受試者需判斷連續出現的兩張臉是否相同。研究結果顯示威廉氏症候群與正常發展控制組的行為結果相同，無歧異表現，但是在腦波表現方面，當正常發展控制組在左右腦清楚區辨特徵改變臉與整體改變臉時，威廉氏症候群無法區辨這兩種臉，且以知覺特徵方式處理臉部整體改變，研究結果以神經生理證據支持威廉氏症候群臉部辨識歧異的認知處理傾向，也提供威廉氏症候群在非語言範疇的中心連貫另一缺損證明。文章也對行為與大腦不對稱表現進行探討。

Purpose: This study investigated the facial processing strategies of Williams syndrome (WS) patients, who exhibit genetic deficits on chromosome 7q11.23. Because of this deficit, this clinical group has been unsuccessful in detecting configural or global information in previous behavioral studies such as the standardized block design test. However, no neurophysiological evidence has been reported regarding this impairment; thus, the event-related potentials (ERPs) technique was used to address this deficit. Methods: Female faces were manipulated (changing the features or configurations) as facial stimuli. The images used to change the features (the eyes or mouth) or configurations were based on other female faces. WS patients (n=13) and their chronological-age matched controls (n=13) participated in this study. The participants assessed the similarities or differences among consecutively presented faces from a set of models, some of which had altered features (feature-changed faces) or configurations (configuration-changed faces). The faces were randomly presented and no duplicates were displayed. Findings: Regarding response latencies and accuracy rates, the behavioral results of WS patients were similar to those of the healthy controls. Both groups demonstrated rapid detection and high accuracy rates when assessing the feature-changed faces, but responded slowly and erred considerably when assessing the configuration-changed faces. However, the groups presented distinct brainwave responses to the configuration-changed faces. The healthy controls processed the configuration-changed faces differently compared with the featurechanged faces in the vertex areas of both hemispheres, whereas the clinical group failed to differentiate these 2 types of facial stimuli. Conclusion: In this study, we discovered neural evidence for a configuration detection deficit among WS patients when processing faces. The results further identified a weak central coherence among WS patients, suggesting a syndrome-general but not syndrome-specific deficit in people with developmental disabilities. Implications: WS patients demonstrated asymmetric brain and behavioral performances during facial processing. This asymmetry was reported in a verbal study that used ERPs and a false memory paradigm. WS patients exhibit genetic deficits that cause atypical development during the early stages of life. These findings were consistent with those of our previous studies pursuing contextual competence, which is defined as the ability to integrate the meanings of words into a contextual theme by using appropriate social knowledge and semantic comprehension; this has been considered a major deficiency among those with autism or right-hemisphere brain damage. Our findings confirmed a deviant central coherence among this clinical group. Neuroconstructivists claim that a small gene mutation during the initial developmental stages can yield devastating effects in long-term development. The deficient configuration detection performance of the WS group provides evidence supporting central coherence deficiency, proving that the interaction between genes and cognition is a dynamic process.