虚假记忆是指对于一个特定的事件，个体将未经历过的经验错误归结为经历过的，或者是记忆中对于该事件的信息表征与原先经历的情形相比有了较大的变化而个体没有觉察到这种变化的一种记忆错误现象。它具有内生性、真伪难辨、高确定性、易感性、固着性、逻辑性等特征，对人们的生活与社会发展都有着重要的影响。由于虚假记忆是难以消除的，并且从行为反应上难以有效和可靠地判定一个记忆是真实的还是虚假的，这就使得如何在神经活动层面上识别虚假记忆、将其与真实记忆分离是当前虚假记忆研究的主要趋势，但是，由于虚假记忆诱发因素的多样性，致使它与真实记忆的神经活动差异在不同的条件下表现不一，因此，要研究虚假记忆与真实记忆神经活动的差异，必须在特定的情境下，通过多种方法、技术的结合方能凑效。情绪是影响记忆与虚假记忆的一个重要因素，情绪刺激一方面能加强记忆效果，另一方面也会增高虚假记忆，由此，情绪刺激对虚假记忆的影响以及在情绪刺激影响下，虚假记忆与真实记忆的神经活动差异是研究两种记忆神经活动分离不可回避的一个问题，同时也是一个影响机制尚不清楚、分离结果不稳定的问题。本研究在对相关文献进行归纳分析、形成上述认识的基础上，设计5个实验开展研究。实验以IAPS图片为刺激材料，采用DRM范式诱发高虚假再认，实验时同步记录和采集行为反应、事件相关电位（ERP）、眼动追踪三方面的数据，实验后使用时程关联分析的方法探讨情绪刺激对虚假记忆的影响，并揭示在情绪刺激情境下，虚假记忆与真实记忆神经活动的差异。实验1考察情绪的效价和唤醒度对虚假记忆的影响，实验结果发现，低唤醒度情绪图片比高唤醒度情绪图片更能诱发虚假记忆，虚假记忆比真实记忆需要更长的加工时间以完成其知觉重现的内生过程。情绪的效价与唤醒度对虚假记忆的影响不一，唤醒度因素对虚假记忆及其与真实记忆神经活动差异的影响时程更早。在再认早期，虚假再认比真实再认需要更深的加工，而到后期则比真实再认需要更多的区域加工，以支持再认时对目标刺激的搜索、比较和辨别。实验2探讨情绪刺激在编码阶段对加工资源的占有情况及其对虚假记忆的影响，实验结果发现，高唤醒情绪刺激在编码时能获取更多的加工资源，强化记忆效果，这种强化效果一方面有利于对真实重现目标刺激的正确辩认，另一方面则成为干扰因素，增加对似是而非诱饵刺激的虚假再认。虚假再认与真实再认的神经活动差异表现在再认阶段早期的知觉编码加工（P1）和晚期的有意回想加工及后提取加工阶段，在高情绪刺激条件下这种差异更显著。早期的差异寓示着真实再认的熟悉效应，后期差异则寓示着虚假记忆比真实记忆需要更多的搜索、比较加工活动。实验3进一步考察强化情绪刺激来源信息记忆对虚假再认的影响，结果发现，来源监测任务影响着两种记忆加工的神经活动差异，使高情绪唤醒刺激虚假记忆与真实记忆加工差异的显著性下降，意味着情绪信息表征对来源信息表征的替代性。两种记忆的神经活动的差异依然表现在再认阶段的早期编码加工和后期的回想和后提取加工，但来源监测任务削弱了这种差异。在记忆信息提取过程中，人们通常会因为各种暗示或提示而产生预警效应，导致记忆结果的各种差异，为此，本研究设计两个实验探讨预警提示对虚假记忆的影响。实验4探讨了一般预警对虚假记忆的影响，结果发现，预警提示产生暗示效应，使被试倾向于采取更为严格和保守的策略对目标刺激进行新、旧辩别，预警提示没有促进被试进行更主动的新旧属性的搜索、提取和比较，而是根据刺激的情绪特征进行更为简要的判别。因此，一般预警会降低情绪刺激的虚假记忆，并使两种记忆的神经活动更为相似。实验5探讨了来源信息监测提示对虚假记忆的影响，结果发现，来源信息监测提示均提高了真实再认和虚假再认的发生率，并降低了加工反应时。这意味着来源信息监测提示对情绪诱发再认偏向的强化作用，来源监测提示强化了再认过程的监测活动和回想活动，这种强化也缩小了虚假再认和真实再认神经活动的差异。根据以上实验结果，本研究认为，虚假记忆加工比真实记忆需要更长的加工时间，情绪效价和唤醒度对虚假记忆都有影响，唤醒度影响更大、发生时间更早。情绪刺激既能提高记忆成绩，也能加剧虚假记忆，原因在于情绪刺激促进了对刺激的熟悉感，当再认目标刺激为编码阶段刺激物重现时，情绪刺激信息表征会促进记忆效果，而当目标刺激与编码阶段相关刺激为似是而非关系时，情绪刺激的信息表征会对再认过程产生干扰作用，提高虚假记忆发生率。情绪条件下虚假记忆与真实记忆神经活动的差异主要表现在再认的早期加工和晚期的回想和后提取加工阶段。来源监测和再认提示对降低情绪刺激的虚假记忆影响有限，反而会使两种记忆神经活动的相似性更大。

False memory is a phenomenon of memory error, which indicates that for a specific event, individuals accept things that never happen, or that compared to the original characterization, individuals’ memory for an event has changed a lot although they are never aware of that. It is characterized as Endogeny, difficult to determine authenticity, high certainty, vulnerability, fixation, longer time for information processing, logic, and so on. False memory has a major impact on society and people’s life and is difficult to eliminate. Moreover, at behavior level, it is hard to determine whether a memory is true or false effectively and reliably, hence how to distinguish false memory from true memory at neuroscience level has become a major direction in the research of false memory. However, because of the diversity of the inducement of false memory, the differences of neural activity between true and false memory differ in different conditions. Therefore, only when multi-method and multi-technique are adopted in specific situations can the research of false memory be successful.Emotion is a key factor to influence memory and false memory for strengthening both of them. So under the condition of emotional stimulus, the difference on neural activity between true memory and false memory can be an interesting question which is not understood well. Based on the summary of related papers, in this study there were 5 experiments which focused on the influence of emotional stimulus on false memory. With IAPS pictures as stimulus material and DRM paradigm inciting high rates of false recognition, the behavioral data, event-related potentials (ERP) and eye tracking were collected in parallel, and the time course association analysis were conducted for their difference.Experiment 1 studied the influence of valence and arousal of emotion on false memory. The results showed that emotional pictures with low arousal incited more false memory than true memory, while false memory cost more processing time to complete the endogenous forming of perception recurrence. Emotional valence and arousal had varied influences, among which the influence of arousal on false memory and on the differences between false and true memory was earlier in the time course. In the early period of recognition, false recognition needed deeper processing, but more areas to compare and distinguish information in the late period than true recognition.Experiment 2 studied how emotional stimulus possessed the processing resource at the encoding stage, and the effect of this possessing on false memory. The results showed that high arousal emotional pictures could intensify the memory with acquiring more processing resources in the encoding period, which was beneficial to the true recognition for the studied pictures but increased the false recognition for the critical lure pictures that not studied. The differences of the neural activity between false and true memory could be seen both at the early perceptive encoding stage (P1) and at the late processing stage invovling recollection and post-retrieval processing, which was more significant for high arousal stimulus than for low arousal stimulus. The differences at the early stage suggested a familiarity effect of true memory, while at the late stage suggested that false memory needed more searching and comparing process than true memory.Source monitoring could also make critical influences on memory performance, so experiment 3 explored the effects of emotional arousal on false memory when enhanced the source monitoring. The results showed that source monitoring made the ERP differences between false and true memory more close. Although there were still the differences of neural activity between false and true memory both at the early encoding stage and the late recollection and post-retrieval stage of recognition, it had been weakened by source monitoring.During the processing of information retrieval, prewarning effect usually arose from suggestion, resulting in the differences of memory results.. Hence, the present study desiged two experiments to investigate how prewarning influenced false memory of emotional stimulus. Experiment 4 investigated the effects of general prewarning on false memory. The results showed the suggestion effects induced by prewarning which made the subjects adopt more conservative and strict strategy for the old/new distinguishing of the emotional stimulus. The prewarning did not promote a more positive searching, retrieving and comparing to the old/new features of emotional stimulus, but a more concise distinguishing according to the emotional features of stimulus. Thus, the general prewarning could reduce the rate of false recognition, and made the neural activity of false and true memory more close. Experiment 5 investigated the influence of prewarning of enhanced source monitoring on false memory. The results showed a faster reaction and increased rates of both true and false recognition when subjects performed the source attribution after being warned to reduce the rate of false alarm while maintain the rate of true recognition to the studied picture. The results suggested that prewarning of source monitoring could not only enhance the emotion-induced recognition bias, but also enhance the monitoring and recollecting of the recognition processing, which, on the other hand, reduced the differences of neural activity between false and true memory.In summary, false memory needed more processing time than true memory, and was influenced by both emotional valence and arousal, in which the influence of arousal was stronger and earlier. Emotional stimulus could improve memory performance, as wells as increasing false memory. This could be attributed to the enhanced familiarity of stimulus promoted by emotion. The target stimulus could promote memory performance when they were the same as the stimulus appeared in the study session, and could induce more false memory for disturbing the processing of mornitoring and evaluating when they are only similar to the related items in the study phase. The differences of neural activity between false memory and true memory appeared in the early and late stage of recognition process. Source monitoring and warning made little influence on inducing false memory rate, furthermore, they increased the similarity of true and false memory.