依戀除了是個體與依戀他人的重要情成連結，也是協助個體去因應威脅的重要情緒調節系統。Bowlby 認為當個體暴露在威脅情境下，依戀系統會被激發，進而影響個體對情緒性訊息的訊息處理歷程，形成特定認知偏誤，而此偏誤與日 後心理過應發展其有關聯性。有鑑於此，本研究將進一步探討依戀特性、情緒性 訊息處理偏誤、及心理過應的關聯性。本研究假設，相較於中性促發情境，個體 在威脅促發情境下，在跨時間歷程上會對情緒性訊息出現先警覺威脅刺激，再警 覺親近刺激之兩階段的注意力偏誤表現。本研究亦假設，其有不同依戀特性的個 體，因採用不同情緒調節策略，對情緒性訊息的注意力偏誤會有組間差異存在，此差異會進一步導致不同心理過應性。本研究從大學普心課程與網路招募共 290 名的研究參與者，依據所有研究參與者在依戀特性的分數，將之分成安全依戀組、疏離依戀組、矛盾依戀組、及懼怕依戀組 組，同時，要求所有研究參與者完成 各項心理過應指標的自陳式作業，以及進行測量注意力警覺反應與注意力抑制反應的電腦化作業。結果發現，相較於中性促發情境，威脅促發情境的確會激發個
體依戀系統的運作，使個體表現出先警覺生氣臉孔，後警覺高興臉孔的兩階段注意力偏誤表現，而此注意力偏誤在其有不同依戀特性的個體存在組間差異，矛盾 依戀組對生氣與高興臉孔都出現警覺-逃避-警覺之矛盾反應，其餘三組則對生氣臉孔表現先警覺後逃避的反應，對高興臉孔表現出逐漸警覺的反應。另外，在安 全依戀組，負向偏誤可顯著預測良好心理過應性，在疏離依戀組與懼怕依戀組，負向抑制反應可顯著預測良好心理過應性，而在矛盾依戀組與懼怕依戀組，正向 偏誤可顯著預測不良心理過應性。綜上，本研究驗證依戀系統對情緒性訊息處理 歷程的影響是決定後續心理過應發展的重要因子。

Attachment is not only an emotional bond between infants and their attachment
figures, but also an important emotion regulation device to manage threat. Bowlby
proposed that encountering with threat may lead to activation of internal attachment
system, in that the processing of emotional information may be influenced and
manifested as cognitive biases. These biases are further associated with later
development of psychological adjustment. Thus, the present study aimed to explore
the associations among attachment, cognitive biases, and psychological adjustment.
We hypothesized that, under threat condition, individuals may display two stages of
attention biases across time course. They will show attention biases towards
threat-related stimuli and followed by towards intimacy-related stimuli. We also
hypothesized that individuals with different attachment characteristics may display
different biases, and these biases may predict psychological adjustment in distinct
ways. The present study recruited 290 participants via Internet advertisement and
course credit requirement at the Department of Psychology, National Taiwan
University. All participants completed a booklet of measures of attachment
characteristics to form secure, dismissive, preoccupied, and fearful attachment groups
and psychological adjustment including depression and anxiety. They then completed
computerized cognitive tasks to assess attention vigilant and inhibition responses
across different time course. We found that, compared to neutral condition, all
particpants displayed two stages of attention biases towards angry faces and followed
by towards happy faces under threat condition. In addition, there were different
responses of cognitive biases in various attachment groups, in which preoccupied
group showed vigilance-avoidance-vigilance pattern towards happy and angry faces,
while other groups showed vigilance-avoidance pattern towards angry faces and
vigilance pattern towards happy faces. Furthermore, negative biases predicted
well-being in secure group and negative inhibition predicted well-being in dismissive
and fearful groups, whereas positive biases predicted ill-being in preoccupied and
fearful groups. Our study demonstrated that the influence of attachment on emotional
processing biases play a critical role in the development of psychological adjustment.