测谎中反馈加工的认知神经机制：一项ERP研究

The Neural Mechanism Processing Feedbacks of Lie Detection:an Event-Related Potential Study

在一个模拟"杀人——侦探"游戏情境中,被试先完成模拟杀人任务,在随后的测谎阶段使用有反馈的ERP测谎范式,以反馈的形式告知被试测谎结果("+0.5"或"-2",分别代表"撒谎逃过检测"或"撒谎被发现")时的脑内时程动态变化。结果发现,在反馈出现后的200～300ms时间段,表示"撒谎被抓住"比"撒谎逃过检测"的反馈信息诱发了一个更正性的ERP成分(P200-300),在300～400ms时间段,两种反馈诱发的ERP成分(P300)没有差异,而在450～500ms时间段,"撒谎逃过检测"比"撒谎被抓住"的反馈信息诱发了一个更负性的ERF成分(N500),负性反馈没有诱发FRN。这些结果表明:P200-300、P300和N500都是与加工测谎结果信息有关的ERY成分。P200-300可能对测谎结果的效价(正性或负性)敏感;P300可能反映被试对测谎结果主观意义大小的评价;而N500可能反映测谎结果所导致的大脑皮层兴奋程度。

Previous research on feedback mainly emphasizes how ERPs relate to Feedback （ such as FRN, P300） and how they change with the changes of feedback~ value, quantity and probability （expected）. Most of the research was conducted in gambling games which were experimental tasks, and the feedback which was in form of figure indicated winning or losing money. Much research investigated the cognitive process preceding a deceptive response, but less attention has been paid to how people perceive the outcomes of deception. In real life, however, it is the outcomes that make people have special mental process when they lie and make their processing of deceptive outcomes different from their general mental processing. In the lie-detection paradigm of this research, positive and negative feedback represent ＂lie-escaped＂ and ＂lie-caught＂ respectively. This kind of feedback may need more complicated cognitive process than the general feedback. We speculate that the cognitive and neural mechanism processing the outcomes of deception may be different from the expectation in usual circumstatnees. In order to explore how people conduct cognitive processing of the outcomes of deception, this research analyzed the ERPs evoked by the feedback in a lie-detection paradigm. Event-related brain potentials were recorded when 17 subjects did this lie-detection task. We mainly analyzed ERPs evoked by the two sorts of feedbacks following deceptive response. The experimental results showed that three kinds of ERPs were evoked by feedbacks following two sorts of deceptive responses, they were P200-300, P300 and N500 respectively, and the negative feedback （＂-2＂） which indicated ＂ lie-caught＂ did not evoke FRN. The feedback, indicating ＂lie-caught＂ , evoked larger amplitude of P200-300 than the feedback indicating ＂lie-escaped＂ did, while the two evoked similar amplitude of P300, which showed that the latter evoked larger amplitude of N500 than the former. The results revealed that, in this research, FRN seemed to reflect instant and motivational evaluation of feedback, and might have no necessary relations with learning process. When processing the feedbacks which indicate the outcomes of deception, P200-300 was distinct from P300. P200- 300 might be sensitive to the consequences of stimulus encoding （positive or negative）, thinking the feedback indicating ＂lie-caught＂ was more negative than the feedback indicating ＂lie-escaped＂ ; while P300 might be sensitive to the subjective significance, thinking the two sorts of feedbacks are similar in subjective significance. P300 represented the process to the subjective significance of feedbacks, and its sensibility to the number of feedback may be due to this fact ： large numbers of feedbacks have more significance of rewards and penalties. When processing the outcomes of lie-detection, N500 might reflect cortical excitability, and ＂lie-escaped＂ might be more exciting than ＂lie-caught＂.