The Selection of Distractors and Eyewitness Identification Accuracy

近年来，社会法学学者对目击证据和证人身份鉴定领域的研究给予了相当大的关注。尽管在从阵容和照片中进行科学鉴定方面取得了重大进展，但所有进展都是一种形式，即在不影响正确鉴定（命中）的可能性的情况下减少错误鉴定的可能性。选择干扰物以满足功能或有效大小的要求一直是减少错误识别率的主要工具。该项目遵循了最近的理论论点，即根据选择干扰物的策略，两个相同功能或有效规模的阵容可能具有不同数量的有利（有益）异质性，从而产生显著不同的命中率。理论和试验数据表明，最大化嫌疑人与干扰物之间差异的实施方法可以提高命中率，从而使证人更有可能正确识别嫌疑人。设计了三个实验，以充分验证异质性思想在不增加错误识别率的情况下提高命中率的可行性。这些实验遵循了一种基本的形式，即先为毫无戒心的目击者上演犯罪，然后向他们展示一组人，并要求他们指认罪犯。将被选择的干扰物(a)与嫌疑人相似，(b)与目击者对罪犯的描述相匹配，或(c)与描述相匹配并使有利方差最大化的干扰物进行比较。在每个实验中，将评估一个置信-准确性关系，期望当有利方差高时，这种关系将更强。本研究对增进对目击证人辨认过程的基本认识具有理论价值。从实际的角度来看，这种关于列队程序动态的知识可以提高在刑事案件中执行司法的质量和效率。最后，在认知心理学领域，本研究将阐明与识别记忆相关的理论问题。

Sociolegal scholars have devoted considerable attention to research in the area of eyewitness evidence and identification in recent years. Although significant advances have been made in the scientific identification from lineups and photospreads, all progress has been of one form, namely reducing the likelihood of false identification without affecting adversely the likelihood of correct identifications (hits). The selection of distracters so as to meet functional or effective-size requirements has been a major tool for reducing false identification rates. This project follows on the recent theoretical argument that, depending on the strategy used to select distracters, two lineups of equal functional or effective size can have different amounts of propitious (beneficial) heterogeneity with the result of having strikingly different hit rates. It is argued from theory and pilot data that implementing ways of maximizing the variance between the suspect and the distractors can improve the hit rates so that witnesses are more likely to correctly identify the suspect. Three experiments are designed to fully test the propitious heterogeneity idea for improving hit rates without inflating false identification rates. The experiments follow a basic format in which crimes are staged for unsuspecting witnesses who later are presented with a lineup and asked to identify the culprit. Comparisons are made between lineups that include distracters who were chosen to (a) resemble the suspect, (b) match the eyewitness's description of the culprit, or (c) match the description plus maximize the propitious variance. A confidence-accuracy relation will be assessed in each experiment with the expectation that the relation will be stronger when propitious variance is high. This research has theoretical value in advancing basic understanding of the process of eyewitness identification. From a practical standpoint, such knowledge about the dynamics of lineup procedures could lead to increased quality and efficiency by which justice is administered in criminal cases. Finally, in the area of cognitive psychology, the study will illuminate theoretical issues associated with recognition memory.