中文摘要：

以情绪面孔为刺激材料，采用改进的内隐知觉驱动测验任务，考察非临床抑郁个体内隐心境一致性记忆的意识加工和无意识加工。结果发现在250ms--350ms时间窗两组被试表现出二者的分离现象，非临床抑郁组的无意识加工过程存在负性心境一致性记忆偏向。在400ms一800ms时间窗内，两组被试不仅加工过程分离，而且在无意识加工过程中均存在负性心境一致性记忆偏向。本研究证实了在内隐心境一致性记忆中存在意识加工，但只有无意识加工存在负偏向且非临床抑郁个体更早表现出该偏向。

英文摘要：

Depression is one of the most prevalent and disabling disorders prone to remission and relapse worldwide. As depression cau- ses heavy social and economic losses, it is important to identify the risk factors related to the onset and maintenance of depressive symp- toms. The cognitive theories of depression believe that the negative bias affects cognitive processes. Beck suggested that depressive schemas favor or facilitate the processing of negative （ mood congruent） stimuli. These mood congruent effects should be observed during automatic processing （ unconscious processing） or at the controlled stage of information processing （ conscious processing）. Mood con- gruent effect is one of the most vulnerable factors for depression. To date, there have been more behavioral studies and fewer cerebral mechanism studies that explore the mood congruent phenomenon and reveal the vulnerable factors for depression. Some studies found that depressed patients remembered more negative faces than positive faces, while the healthy individuals demonstrated an opposite pat- tern with positive bias for positive faces. However, a number of studies have failed to observe the significant mood-congruent memory bi- as. The inconsistency in mood congruent memory phenomenon may be caused by the confusion of the unconscious elaboration with the conscious processing underlying memory test. This study aims to discriminate the awareness and unawareness processing of mood-con- gruent memory in the transient sad mood state. Seventeen sub-clinical depressed and seventeen matched healthy individuals performed the improved perceptual-driven paradigm which contained awareness and unawareness processing while event-related potentials （ERPs） were recorded after the sad-mood induction. The results revealed main effects on study style, significant interaction effect between study style and valence, further analysis showed valence occurred in studied-brief and differed between two groups during the 250ms-350ms, indicating smaller mean ar