根據Koster、De Lissnyder、Derakshan及De Raedt（2011）的注意力轉移困難假說，憂鬱者因注意力控制能力受損，而難以從負向訊息中轉移注意力，進而引發反芻，並再度強化憂鬱症狀。雖然反芻可再細分為深思反省與苦惱自責式反芻，但Koster等人未探究注意力轉移困難對苦惱自責式反芻的影響。此外，過往注意力訓練研究作業眾多且效果不一，又偏重改善個體注意力投入以減緩憂鬱。然而，卻鮮少探討注意力訓練對注意力轉移困難的介入，能否改善個體的苦惱自責式反芻程度。因此，本研究將同時探討注意力訓練能否改變高苦惱自責式反芻者對負向訊息的注意力偏誤（含注意力投入與轉移困難），進而降低苦惱自責式反芻程度。本研究以點偵測作業為注意力訓練作業，將高、低苦惱自責式反芻者隨機分派至注意力訓練組或注意力訓練控制組，接受為期兩週、共四次的注意力訓練，並於前、後測階段注意力測量作業中，檢驗對負向訊息的注意力偏誤與三階段中苦惱自責式反芻程度。本研究結果發現高苦惱自責式反芻者對負向訊息無明顯注意力偏誤，注意力訓練作業對高苦惱自責式反芻者的注意力歷程未有明顯影響，乃至苦惱自責式反芻程度的時序變化與接受注意力訓練與否無明顯關聯。本研究結果不支持原先假設、注意力困難假說及過去研究結果。然而，過往學者多強調個體高度負向認知與憂鬱情緒對注意力偏誤的影響，故本研究事後同時納入憂鬱與苦惱自責式反芻程度，欲探討憂鬱苦惱自責式反芻者對負向訊息有無注意力偏誤，乃至注意力訓練對憂鬱苦惱自責式反芻者注意力偏誤的訓練效果。本研究結果僅發現在修正版Posner作業中，憂鬱苦惱自責式反芻者更容易將注意力從負向訊息中轉移開來；在點偵測作業中，未有組間效果；在注意力訓練中，未有訓練效果。最後，本研究將於討論中，探討研究結果的可能原因，並提出本研究限制與未來研究上的建議。 / According to the impaired disengagement hypothesis (Koster, De Lissnyder, Derakshan, & De Raedt, 2011), the dysphoric that are difficult to disengage from negative stimuli due to low attentional control tend to ruminate, and then even worsen their depressive symptoms. Actually, rumination can be differentiated into two components: reflective pondering and brooding, but the core tenet of impaired disengagement hypothesis only puts the emphasis on rumination rather than brooding. Besides, there are many studies investigating the attention-training effects on depressive symptoms rather than rumination. To date, no studies even have investigated the training effects on impaired disengagement and brooding. Therefore, we aimed to examine the effects of attention training on attention bias toward negative stimuli, impaired disengagement from negative stimuli and brooding level in brooders. We investigated the training effect in brooding and non-psychiatric control participants via dot-probe task. During a two-week period, all of the participants were randomly assigned to complete 4 sessions of either attention training or no training. Also, participants completed two attentional tasks examined attention bias at baseline and post-training, and self-reported questionnaires of brooding and depressive symptoms at baseline, post-training, and follow-up. Overall, results indicate that brooders didn’t show attention bias to negative stimuli. Also, no beneficial effects of attention training on attention bias and brooding level were found in brooders. The previous hypothesis, impaired disengagement hypothesis and studies in the past were not supported. However, many cognitive models of depression have postulated that individuals with high levels of negative cognition and depressive affect tend to maintain their attention toward negative information. Therefore, we took levels of depression and brooding into account, and aimed to examine the effects of attention training on both attention bias toward negative stimuli and impaired disengagement from negative stimuli in depressive brooders. Results indicate that depressive brooders tend to disengage from negative stimuli in modified Posner task. No other findings in dot-probe task and attention-training task. Implications of these findings in depressive brooders are discussed and directions for future research are advanced.