Video-recorded Interrogations: Beyond Camera Perspective

摘要在过去的二十年里，许多被揭露的错误定罪都源于刑事调查的审讯阶段，有时会从被拘留的犯罪嫌疑人那里获得胁迫或虚假的供词。一个被广泛推荐的帮助防止这一特殊问题的解决方案是对拘留审讯进行全程录像。这种程序上的修改确实有可能改进司法工作;但是，谨慎起见，需要认真考虑其可能的缺陷。在基础心理科学的指导下，最著名的是关于虚幻因果关系的文献，G。丹尼尔·拉斯特（Daniel Lassiter）25年多的系统研究经验证明了录像审讯中存在着现在通常所说的摄像机视角偏差。也就是说，摄像机将观察者的视觉注意力引导到嫌疑人身上（而不是审讯者或嫌疑人和审讯者）的视频往往会对嫌疑人产生更有偏见的评价，他们的陈述被评估为更自愿，他们更有可能被判有罪，他们收到更严厉的判决建议。重要的是，Lassiter博士的研究项目还证明了与视频记录相关的优势;例如，如果摄像机采用等焦视角，则可以向事实调查人员提供更详细的审讯记录，而不会引入偏见或破坏准确的评估。虽然很明显，摄像机视角是任何旨在保护无辜者的录像政策都应该解决的重要问题，但不幸的是，这并不是行为科学文献提出的唯一问题。拟议中的研究借鉴了心理科学的四个额外领域，以进一步得出视频记录审讯的潜在局限性。一系列的实验计划，将1）确定这些假设的限制可能会在多大程度上影响实际的事实发现者的视频记录的审讯评价，和2）评估可能的策略，成功地抵消他们。最初的一组研究调查了这样一种可能性，即在某些情况下，即使是同等聚焦的视频记录也可能无法完全保护少数民族嫌疑人，这是由于Lassiter博士及其同事所说的种族显着性偏见。第二组检查潜在的判断毅力，辩方必须克服，以允许事实调查人员正确地重新解释视频审讯，控方凭借审判惯例，有机会首先确定被告的罪行。第三类侧重于事实调查者可能犯下的来源监测错误，这些错误可能妨碍他们准确得出以下结论的能力：录像的供词叙述实际上受到警方在审讯初期提供的信息的污染。最后一组借鉴了conconciliation水平理论，以测试是否诱导一个抽象的（高）或具体的（低）水平的精神conciliation之前，观看视频记录的审讯促进更准确的评估供词的可靠性。 拟议的工作还将为法律的从业人员和政策制定者提供明确的基于研究的指导方针，这将进一步提高录像改革履行其承诺的可能性，即最大限度地减少无辜者因审讯诱导的虚假供词而被定罪的可能性。在Lassiter博士以前资助的NSF项目中，几名研究生和本科生能够直接了解基本心理学理论和研究如何引起对司法系统乃至整个社会具有相当实际意义的科学调查。同样，拟议的工作将使新一代学生和未来的学者接触到一个纲领性研究议程的所有阶段，该议程同时具有高度的科学严谨性和即时的现实世界适用性。

AbstractMany of the wrongful convictions that have been brought to light over the last two decades have their roots in the interrogation phase of criminal investigations where coerced or false confessions are sometimes extracted from detained crime suspects. A solution that has been widely recommended to help prevent this particular problem is to video record custodial interrogations in their entirety. Such a procedural modification indeed has the potential to improve the administration of justice; however, prudence calls for careful consideration of its possible pitfalls. Guided by basic psychological science, most notably the literature on illusory causation, Dr. G. Daniel Lassiter's more than 25 years of systematic investigation has empirically documented the existence of what is now commonly known as the camera perspective bias in video-recorded interrogations. That is, videos in which the camera directs observers visual attention onto suspects (rather than onto interrogators or equally onto suspects and interrogators) tend to produce more prejudicial evaluations of suspects, their statements are assessed as more voluntary, they are more likely to be judged guilty, and they receive more severe sentence recommendations. Importantly, Dr. Lassiter's research program also demonstrated that the advantages associated with video recording; for example, a more detailed record of the interrogation is provided to fact finders without introducing bias or undermining accurate assessments if an equal-focus perspective is taken by the video camera. Although it is clear that camera perspective is an important issue that any video-recording policy designed to protect the innocent should address, it is unfortunately not the only concern suggested by the behavioral science literature. The proposed research draws on four additional areas of psychological science to derive further potential limitations of video-recorded interrogations. A series of experiments are planned that will 1) determine the extent to which these hypothesized limitations would likely detrimentally influence actual fact finder evaluations of video-recorded interrogations, and 2) evaluate possible strategies for successfully counteracting them. An initial set of studies investigates the possibility that even an equal-focus video recording in certain circumstances may not fully protect minority suspects, due to what Dr. Lassiter and colleagues call a racial salience bias. A second set examines the potential judgment perseverance the defense must overcome to permit fact finders to correctly reinterpret a video interrogation that the prosecution, by virtue of trial convention, has the opportunity to first characterize as establishing a defendant's guilt. A third set focuses on the source monitoring errors fact finders may commit that could impede their ability to accurately conclude that a video-recorded confession narrative was actually contaminated by information provided by police earlier in the interrogation. A final set draws on construal level theory to test whether inducing an abstract (high) or concrete (low) level of mental construal prior to viewing video-recorded interrogations promotes more accurate assessments of a confession's reliability. The proposed work will also yield clear research-based guidelines for legal practitioners and policy makers that will further improve the likelihood that the video-recording reform will fulfill its promise to minimize the possibility that an innocent person will be convicted based on an interrogation-induced false confession. In Dr. Lassiter's previously funded NSF projects, several graduate and undergraduate students were able to learn firsthand how basic psychological theory and research can give rise to scientific investigations that are of considerable practical import to the justice system and thus society more generally. Similarly, the proposed work will expose a new generation of students and future scholars to all phases of a programmatic research agenda that is simultaneously high in scientific rigor and immediate real-world applicability.