No fallacies in the formulation of the paternity index.

摘要: Abstract In a recent publication, Li and Chakravarti claim to have shown that the paternity index is not likelihood ratio. They present method of estimating prior probability from sample previous court cases on basis exclusions nonexclusions. propose calculating posterior this estimated test result expressed as exclusion/nonexclusion. Their wrong—the likelihood-ratio, is, ratio observation conditional two mutually exclusive hypotheses. Their proposed has been long known, applied several samples, inferior (in terms variance estimate) maximum estimation based all phenotypic information available. “new method” use less informative 1/(1 – PE) instead Gurtler's fully X/Y (Acta Med Leg Soc Liege 9:83–93, 1956), but otherwise indentical Bayesian approach originally introduced by Essen-Moller in 1938.