Department of Anthropology, University of West Florida, Pensacola, FL, USA.
SNA International, Alexandria, VA, USA.
J Forensic Sci. 2021 Sep;66(5):1617-1626. doi: 10.1111/1556-4029.14761. Epub 2021 Jun 28.
Due to their medicolegal repercussions, forensic anthropology conclusions must be reliable, consistent, and minimally compromised by bias. Yet, a synthetic analysis of the reliability and biasability of the discipline's methods has not yet been conducted. To do so, this study utilized Dror's (2016) hierarchy of expert performance (HEP), an eight-level model aimed at examining intra- and inter-expert reliability and biasability (the potential for cognitive bias) within the literature of forensic science disciplines. A systematic review of the forensic anthropology literature was conducted (1972-present), including papers published in the American Journal of Physical Anthropology, Forensic Anthropology, Forensic Science International, and the Journal of Forensic Sciences and Anthropology Section abstracts published in the Proceedings of the Annual Meeting of the AAFS which matched keywords such as "forensic anthropology," "bias," "reliability," "cognition," "cognitive," or "error." The resulting forensic anthropology HEP showcases areas that have ample research and areas where more research can be conducted. Specifically, statistically significant increases in reliability (p < 0.001) and biasability (p < 0.001) publications were found since 2009 (publication of the NAS report). Extensive research examined the reliability of forensic anthropological observations and conclusions (n = 744 publications). However, minimal research investigated the biasability of forensic anthropological observations and conclusions (n = 20 publications). Notably, while several studies demonstrated the biasing effect of extraneous information on anthropological morphological assessments, there was no research into these effects on anthropological metric assessments. The findings revealed by the forensic anthropology HEP can help to guide future research, ultimately informing the development and refinement of best-practice standards for the discipline.
由于其法医学上的影响，法医人类学的结论必须是可靠的、一致的，并且最小程度地受到偏见的影响。然而，目前还没有对该学科方法的可靠性和偏差进行综合分析。为此，本研究利用了 Dror（2016）的专家绩效层次结构（HEP），这是一个八级模型，旨在检查法医学学科文献中内部和专家之间的可靠性和偏差性（认知偏差的可能性）。对法医人类学文献进行了系统的回顾（1972 年至今），包括发表在《美国体质人类学杂志》、《法医人类学》、《法医学国际》和《法庭科学杂志》和美国法医学协会会议记录中的论文摘要，这些论文都与“法医人类学”、“偏见”、“可靠性”、“认知”、“认知”或“错误”等关键词匹配。由此产生的法医人类学 HEP 展示了有大量研究的领域和可以进行更多研究的领域。具体来说，自 2009 年（NAS 报告发布）以来，发现可靠性（p<0.001）和偏差性（p<0.001）出版物显著增加。大量研究检验了法医人类学观察和结论的可靠性（n=744 篇出版物）。然而，只有很少的研究调查了法医人类学观察和结论的偏差性（n=20 篇出版物）。值得注意的是，虽然有几项研究表明外来信息对人类学形态评估有偏倚作用，但没有研究这些影响对人类学度量评估的影响。HEP 揭示的法医人类学发现可以帮助指导未来的研究，最终为该学科的最佳实践标准的制定和完善提供信息。
