Fallis, Don; Liddell, Gerrard - In: Statistics & Probability Letters 60 (2002) 2, pp. 169-182
Goldman and Shaked (Statist. Probab. Lett. 12 (1991) 415) show that, for all reciprocally convex measures of truth possession, experiments are always "objectively" expected to increase a scientist's degree of truth possession. We show that this result is optimal. Further, we argue that all...