Minakata K, Asano M
Z Rechtsmed. 1980;85(3):189-97. doi: 10.1007/BF02116319.
The relation between the exclusion probability (E) and the paternity probability is derived by assuming the distributions of logarithm of paternity likelihood ratio, log(Y/X) for true fathers and unexcluded non-fathers as the normal distributions. Under this assumption the value log(1-E) is equal to the mean of the mean value for true fathers (a) and that for unexcluded non-fathers (b), i.e., log(1-E)=(a+b)/2. This relation holds quite well for the various actual distributions of log(Y/X) of father-child combinations and those of father-mother-child combinations using 14 blood group systems. Therefore, the derived relation is found to be a convenient way to deduce one of the three quantities (E, a, b) from the remaining two quantities in the actual distributions.
通过假设真实父亲和未被排除的非父亲的父权似然比的对数log(Y/X)的分布为正态分布，推导出排除概率（E）与父权概率之间的关系。在此假设下，值log(1 - E)等于真实父亲的均值（a）与未被排除的非父亲的均值（b）之和的均值，即log(1 - E) = (a + b)/2。对于使用14种血型系统的父子组合以及父母子组合的log(Y/X)的各种实际分布，该关系都非常适用。因此，发现所推导的关系是从实际分布中的其余两个量推导出三个量（E、a、b）之一的便捷方法。
