Attempt at an estimation of the utility of paternity probabilities, calculated from blood groups

Abstract

From a comparison of fatherhood probability distinctions, established by Bayes’ formula from 6 blood group sets — 185: mother, father, child tercets and 146 mother, child, randomly chosen man tercets (on omission of exclusions in the latter tercets) an attempt to assess the utility of those probabilities in the estimation of kinship in causes of fatherhood origin investigation. Very high estimates of fatherhood probabilities as well as very low ones occurred in both materials examined, but a well determined percentage predominance of real families over „false families” was found in the probability range 90%—99%. A predominance of „false families” over real families occurred below 55% of fatherhood probability. In the probability ranges 55% — 90% the error in the relationship estimate on the ground of serologic characteristics is too large to be reliable for sentencing by court in cases of controversial paternity. A comparison of paternity probabilities on the ground of anthropological researches including 180 morphological characteristics with the probability values from blood group in 123 expert evidences led in many cases to contradictory results. In expert evidences with two putative fathers, the results of paternity probability from the blood group in most cases failed to show distinctly to which of two tested men the paternity should be more likely attributed. The utility of paternity probability estimates on the basis of only serological structures is very small in the judicature.