| Abstract: | A "realistic" prior probability is always based on case experience (Akten-a-priori). In serological opinions pertaining to parentage, the realistic prior probability is only one piece of information in the whole body of evidence before the judge and does not have any special significance per se. There is no such thing as a "neutral" prior probability. It either implies "ignorance," in which case it cannot be "information," or it must be taken in connection with the utility principle, in which case it is not a "probability." The utility principle is defined in law and cannot be expressed in figures. The utility principle takes effect only when the judge reaches a decision (on the basis of all the evidence before him). It determines the relative importance of the participant's objects of legal protection which are at issue in the case. The expert is bound to apply a neutral utility component, i.e., in a two-hypothesis case (the normal situation) the significance of both the null and the counter hypothesis must carry the same weight. A null and/or a counter hypothesis can combine several single hypotheses; the mean value of their frequencies is taken. As a rule, one should avoid using a "prior case probability" ("Akten-a-priori") when calculating a W value. An "expectation of error" should be as realistic as possible and hence be obtained using a "prior case probability." |
