| 摘要: | 異族效應（ORE）是一種穩定、強大並且跨文化的現象，指稱的是相較於辨識其他種族的臉孔，人們在識別與自己同種族的臉孔更為容易，而此效應被認為是由於對其他種族臉孔的經驗較少而產生的。然而，關於在判斷臉孔吸引力時是否會觀察到異族效應，還有待確定。過往研究指出熟悉度會影響對人臉的記憶和對人臉的吸引力程度判斷，但臉孔吸引力和熟悉度間的確切關係目前尚未有定論。
在本研究的實驗一中，受試者的記憶表現顯示出清晰的異族效應：在進行劍橋臉孔記憶測試作業（CFMT）時，高加索受試者對澳大利亞裔的臉孔比華裔的臉孔有更佳的記憶，東亞(台灣)受試者對華裔的臉孔比澳大利亞裔的臉孔有更佳的記憶，而南亞受試者則對兩個族裔的臉孔有相似的記憶表現。另一方面，三組受試者皆將澳大利亞裔的臉孔評為比華裔的臉孔更具吸引力。本實驗也利用種族接觸問卷（RCQ）衡量三組受試者與高加索人和華人的社會互動情況，但未發現有任何證據顯示受試者對不同族裔臉孔的記憶準確性和吸引力判斷，與他們在RCQ中的填答反應有相關。
為了檢視對同一張臉的吸引力判斷和再認記憶的表現，是否會展現出和實驗一相同的結果，我們在實驗二中檢視東亞(台灣)受試者對於相同臉孔的吸引力評分與再認記憶表現。本實驗的結果與實驗一結果一致，顯示受試者判斷澳大利亞裔臉孔比華裔臉孔更具吸引力。另一方面，在再認記憶作業中的表現則顯示：與其他種族（澳大利亞）臉孔相比，受試者對自己的種族（華人）的臉孔有更好的識別記憶，其結果反映在對刺激材料有較高的敏感度和較保守的辨識標準上。此外，我們將受試者對每張臉孔的吸引力評分與對該張臉孔的再認記憶表現進行相關分析，結果顯示受試者感知到的臉孔吸引力與相對應的記憶敏感度和判斷標準呈負相關；具體來說，吸引力低的臉孔比起吸引力高的臉孔更容易被人們記住。
在實驗三中，我們在東亞(台灣)受試者進行臉孔吸引力判斷和再認記憶作業時，同步記錄事件相關電位（ERPs）的神經生理訊號，比較對自身與其他種族、具吸引力與缺乏吸引力的男性/女性臉孔的大腦反應。對本實驗腦電波訊號的分析顯示：在兩個作業中，位於額葉、中央區電極記錄到的N1（60-80 ms），以及在頂葉、頂枕葉與枕葉區電極記錄到的P1（80-120 ms）、N170 （140-200 ms）和 LPC（500-700 ms）等事件相關電位成份，在處理臉孔時均有顯著的增強。在吸引力判斷作業中，自己種族的臉孔比起其他種族的臉孔引發更大的P1振幅；而在所有早期（N1、P1、N170）和晚期（LPC）的ERP成份中，由具吸引力的臉孔和不具吸引力的臉孔所引發的反應都沒有顯著差異。在再認記憶作業之"命中"回應(即正確辨認出有出現過的臉孔)的腦電波訊號中，其它種族的臉孔比起自己種族的臉孔引發更大的LPC振幅，但沒有偵測到臉孔吸引力的效果。
綜上所述，目前本研究的結果顯示，只有與種族有關的臉孔特徵、而非和臉孔吸引力有關的特徵，其處理才會直接受到對臉孔之熟悉度和經驗的調節。欲釐清臉孔吸引力、熟悉度和記憶在行為和神經生理層次的交互作用，則需要進一步的研究。
;The other-race effect (ORE) is a robust and cross-cultural phenomenon that shows people recognizing own-race faces better than other-race faces. The ORE is assumed to result from less experience with other-race than own-race faces. However, it remains to be determined whether the ORE would be observed in appreciating facial attractiveness. Previous research has shown that familiarity would affect attractiveness judgments and memory of faces, but the exact relationship between facial attractiveness and familiarity is still debated.
In Experiment 1 of the current study, participants’ performance in the Cambridge Face Memory Test (CFMT) showed a clear ORE in memory, as Caucasian participants remembered Australian faces better than Chinese faces, East-Asian (Taiwanese) participants remembered Chinese faces better than Australian faces, while South-Asian participants remembered faces of both races equally. On the other hand, Australian faces were rated as more attractive than Chinese faces by the participants from all three groups. Additionally, a Race Contact Questionnaire (RCQ) was administered to the participants to measure their social interaction with Caucasian and Chinese people. We did not find any evidence that participants’ memory accuracy and attractiveness judgment correlated with their answers in the RCQ.
To investigate the effect of facial attractiveness on recognition memory of the same faces, and whether participants are differentially sensitive to own-race and attractive faces, we collected attractiveness ratings and measured recognition memory performance on the same faces from East-Asian (Taiwanese) observers in Experiment 2. The results replicated the findings of attractiveness ratings in Experiment 1 by showing that Australian faces were rated as more attractive than Chinese faces. On the other hand, better recognition memory (as reflected in higher sensitivity and more conservative criterions) was found for own-race (Chinese) than other-race (Australian) faces. Further, an item-wise analysis was performed by correlating the mean attractiveness ratings of individual faces with the mean memory performance. The results revealed that perceived facial attractiveness negatively correlated with corresponding memory sensitivity and criterions. Specifically, attractive faces were remembered worse than unattractive faces.
In Experiment 3, event-related potentials (ERPs) were recorded simultaneously during the facial attractiveness judgment and the face recognition memory tasks, and ERP responses to own- vs. other-race and to attractive vs. unattractive male-female faces were compared. The analysis of the neurophysiological signals showed an enhanced N1 (60-80 ms) component at frontocentral sites and the P1 (80-120 ms), N170 (140-200 ms) and LPC (500-700 ms) components at parietal, parieto-occipital and occipital sites during face processing in both tasks. During the face attractiveness judgment task, a stronger P1 component was elicited by own- than other-race faces. No significant difference was observed at the early (N1, P1, N170) or late (LPC) ERP components for attractive vs. unattractive faces. During the face recognition memory task, a larger LPC component associated with the “hits” responses (i.e., correct identification of seen faces) to other-race than own-race faces was observed, while no effect of attractiveness was detected.
In summary, findings of the current study suggest that only the processing of facial characteristics related to race but not to attractiveness is directly modulated by familiarity and experience with faces. Further research is needed to clarify the behavioral and neurophysiological interactions among attractiveness, familiarity, and memory. |
