Balding D J, Donnelly P
School of Mathematical Sciences, Queen Mary and Westfield College, London, England.
J Forensic Sci. 1996 Jul;41(4):603-7.
The paper is concerned with the strength of DNA evidence when a suspect is identified via a search through a database of the DNA profiles of known individuals. Consideration of the appropriate likelihood ratio shows that in this setting the DNA evidence is (slightly) stronger than when a suspect is identified by other means, subsequently profiled, and found to match. The recommendation of the 1992 report of the US National Research Council that DNA evidence that is used to identify the suspect should not be presented at trial thus seems unnecessarily conservative. The widely held view that DNA evidence is weaker when it results from a database search seems to be based on a rationale that leads to absurd conclusions in some examples. Moreover, this view is inconsistent with the principle, which enjoys substantial support, that evidential weight should be measured by likelihood ratios. The strength of DNA evidence is shown also to be slightly increased for other forms of search procedure. While the DNA evidence is stronger after a database search, the overall case against the suspect may not be, and the problems of incorporating the DNA with the non-DNA evidence can be particularly important in such cases.
本文关注的是当通过在已知个体的DNA图谱数据库中进行搜索来识别嫌疑人时，DNA证据的强度。对适当似然比的考量表明，在这种情况下，DNA证据比通过其他方式识别嫌疑人、随后进行图谱分析并发现匹配时（略）更强。美国国家研究委员会1992年报告中关于用于识别嫌疑人的DNA证据不应在审判中出示的建议，因此似乎过于保守。普遍认为当DNA证据来自数据库搜索时其效力较弱，这一观点似乎基于一种在某些例子中会导致荒谬结论的原理。此外，这种观点与得到大量支持的证据权重应由似然比衡量的原则不一致。对于其他形式的搜索程序，DNA证据的强度也显示出略有增加。虽然经过数据库搜索后DNA证据更强，但针对嫌疑人的整体案件情况可能并非如此，并且在这种情况下将DNA证据与非DNA证据相结合的问题可能尤为重要。
