情緒低落者在負向情緒誘發後的情緒一致/不一致記憶：情緒改變量之角色

Abstract

本研究探討在負向情緒誘發後個體的情緒一致或不一致的記憶可近性，並以特質情緒和情緒改變量為切入點，討論情緒低落者使用情緒不一致回憶來改變其情緒的現象。MacLeod與Campbell（1992）的研究顯示情緒誘發可選擇性地促進與情緒一致的事件的記憶可近性；然而，無論是當下處在快樂或不快樂情緒的個體，快樂事件的回憶的反應時間比不快樂事件的回憶的反應時間還要快，此研究結果指出了在負向情緒中的情緒不一致記憶的效應。本研究假設記憶的選擇可近性之表現與個體的憂鬱程度、和負向情緒誘發之後的情緒改變量有關。實驗設計為根據貝氏憂鬱量表分數將八十七名台灣大學學生受試者分為高、低憂鬱組；經過負向情緒誘發作業後，隨機地給予不同的情緒測量指導語操弄（告知或無告知受試者其情緒會被測量），之後進行隨機提供正負向線索詞（各40詞）之回憶作業，電腦記錄其反應時間。將受試者依照其情緒改變量分為高、低改變組，進行2 （憂鬱組別：高、低） × 2 （情緒改變: 高、低） × 2 （線索詞價性：正、負），反應時間為依變項之重複量數變異數分析，與2 （憂鬱組別：高、低） × 2 （情緒改變: 高、低） × 3 （測量時間：負向情緒誘發前、負向情緒誘發後、線索回憶作業後），整體負向情緒狀態分數為依變項之重複量數變異數分析。結果顯示，憂鬱組別、情緒改變、線索詞價性有顯著的三階交互作用，低憂鬱－低情緒改變者有明顯的情緒不一致效應，低憂鬱－高情緒改變者、高憂鬱－高情緒改變者、和高憂鬱－低情緒改變者雖沒有明顯的情緒一致或不一致效應，但前兩者在平均反應時間上對正向記憶快於對負向記憶，後者卻對負向記憶之反應時間快於對正向記憶之反應時間。在情緒改變方面，低憂鬱者以情緒不一致記憶來修復情緒的效能，較高憂鬱者佳。本研究結果驗證了過去研究者對記憶可近性在不同情緒上的不對稱性的假設，並在心理治療上，提供以情緒改變量為切入點來探討情緒修復策略之實徵研究證據。

This study investigated the effects of negative mood on recall latencies for valence of cues. MacLeod and Campbell (1992) found mood-induction selectively facilitated memory accessibility for mood-congruent events. Nonetheless, reduced recall latencies of happy events, rather than of unhappy events, was also found in both happy and unhappy mood conditions, indicating the involvement of mood-incongruent memory. The present study hypothesized that selective accessibility of memory might be associated with the interplay of the individual difference in degree of dysphoria and mood change amount after negative mood induction. Eighty seven college students were grouped as high/low-dysphoria according to their BDI scores. After negative mood induction, they undertook a memory accessibility task to recall their autobiographic memories primed by 40 positive cues and 40 negative cues, with the order of the cues and cue valence counterbalanced. A 2 (dysphoric group: high vs. low) × 2 (mood change amount: small vs. large) × 3 (time: pre- , post-mood-induction, and post-recall-task) × 2 (cue valence: negative vs. positive) repeated measures ANOVA was conducted on response latency. The result showed a significant dysphoria × degree of mood change × cue valence interaction effect in addition to significant two way interaction effects. Specifically, the high dysphoria with small amount of mood change showed a mood-congruent memory accessibility effect, while the other three groups showed a mood-incongruence effect. Furthermore, it was noted that for high dysphoric individuals while memory accessibility for negative memories was not associated with the amount of mood change, for positive cues, the large change group did show a reduced response latency, while increased response latency was found for the small change group. The implications of the results in terms of cue valence, amount of mood change, dysphoria vs. non-dysphoric, and effect of memory accessibility on mood repair process are discussed and future research possibilities are proposed.