Hummel K, Conradt J
Z Rechtsmed. 1978 Jul 27;81(3):217-22. doi: 10.1007/BF00201270.
More than 1800 HLA-typed mother-child-putative father triplets were biostatistically analyzed on the basis of the Essen-Möller principle. 14 of them had W-values (= probability of paternity) of less than 50%, the lowest value being 4,4%. Using a specially designed computer program, exclusion probabilities between 68 and 96% were calculated. This demonstrates (a) that nonexclusion is not necessarily in every case a positive indication of the paternity of the man involved; and (b) that the probability of exclusion as a biological method of establishing paternity is not a suitable piece of evidence. The W-value (= "probability of paternity"), obtained on the basis of the Essen-Möller principle, provides complete information, and is alone conclusive.--The high proportion of non-fullhouse children among the 14 cases allows the conclusion that the more comprehensive the specificity spectrum of the antiserums used in the typing is, the fewer the number of such cases.
基于埃森 - 默勒原理，对1800多个进行了HLA分型的母子 - 推定父亲三联体进行了生物统计学分析。其中14个三联体的W值（=父权概率）小于50%，最低值为4.4%。使用专门设计的计算机程序，计算出的排除概率在68%至96%之间。这表明：（a）不排除在每种情况下不一定表明所涉男子具有父权；（b）作为确定父权的生物学方法，排除概率不是合适的证据。基于埃森 - 默勒原理获得的W值（=“父权概率”）提供了完整信息，且仅凭此即可得出结论。——在这14个案例中，非完全匹配孩子的比例很高，由此可以得出结论：分型所用抗血清的特异性谱越全面，此类案例的数量就越少。
