Abstract

The passage criticized by Mr. Linhart as “misleading” may be clarified as follows. Linhart is quite right that a method of interval estimation including a formula equivalent to my VI may be based on inverse probability, and that probability values considerably greater than zero may be thus obtained. The method of inverse probability to which I refer in the criticized passage, however, is that of Carnap, according to which the inverse probability of a law on the basis of finite evidence is zero. My statement that “... our method of interval estimation is superior to that of inverse probability, in that the probability values yielded by the former are considerably greater than those given by the latter...” remains true, since it is Carnap's method, which is not an interval method, of which I am speaking.