Abstract: Abstract: With the purpose of reviewing current researches on false confession, this paper introduces the typology of false confession, analyses factors that may lead to false confession in terms of personal aspect and situational aspect, and presents computer-crash paradigm and novel experimental paradigm for eliciting false confessions in laboratory. Researchers outlined false confessions to 3 types: voluntary, coerced-complaint, and coerced-internalized. This descriptive typology provides a frame for people to understand false confession systematically. The personal aspect of false confession concerns factors relevant to suspects, including compliance, suggestibility and age, with questionnaires as the dominant method. Researches show that innocent people who are high in compliance and suggestibility and low in age may be risky of giving false confession. As for the situational aspect, false incriminating evidence and “minimization” are the two key factors to which attentions have been paid by the researchers, with laboratory experiments as the dominant method. The computer-crash paradigm devised by Kassin and Kiechel (1996) is widely used in experiments for eliciting false confessions in laboratory with false incriminating evidence. Pace (slow or fast) and evidence (witness or no witness) are the two independent variables in this experiment. Subjects, who were accused of damaging a computer by pressing the wrong key which was prohibited to press, denied the charge initially. But some of them turned out to confess after the confederate provided a witness of seeing or not seeing the subject pressed the wrong key. Confession rate, which was lowest in the slow-pace*no witness group and highest in the fast-pace*witness group, was significantly different between the two groups. The result of this experiment indicates that it is possible for people to internalize guilt for a crime they did not committed and that the presentation of false confession may increase the risk. The novel experimental paradigm, proposed by Russano et al. (2005), measures the behavioral effects of minimization in interrogations. Minimization, a tactic used widely by interrogators, was introduced in this paradigm as an independent variable. Paired with a confederate in a problem-solving study, subjects were instructed to work alone on some problems and together with a confederate on other problems. The paradigm involves guilty and innocent participants being accused of violating the experimental prohibition. The result indicates that the use of minimization increased the rate of both true and false confessions. Researches on false confession are discussed finally from perspectives of operational criterion, case study and ecological validity. Further research should provide operational criterion to describe false confessions more specifically so that interrogators would be aware of the conditions that they should keep away from. Also, researchers should pay more attention on case study to broaden and deepen the understanding of false confession. Then, several problems in the computer-crash paradigm need to be solved so as to improve the ecological validity of the experiment. Finally, the paper give some suggestions on the amendment of the criminal procedure law from the practice.

Key words: false confession, suspect, personal aspect, situational aspect, suggestion

摘要： 本文以虚假供述的研究方法为起点，结合案例，对自愿型、强迫—服从型和强迫—内化型三种类型的虚假供述进行介绍，从个人因素与情境因素两个方面，对虚假供述的影响因素加以分析。其中，着重阐述了虚假供述的实验室研究范式。在评析影响因素的基础上对我国刑事诉讼的修改提出立法建议，以期在法律上控制影响因素的作用。

关键词: 虚假供述 , 嫌疑人, 个人因素, 情境因素, 建议