5
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: not found

      Equal prior probabilities: can one do any better?

      Forensic Science International
      Bayes Theorem, Expert Testimony, Forensic Sciences, methods, Humans, Likelihood Functions, Models, Theoretical, Probability Theory, Shoes

      Read this article at

      ScienceOpenPublisherPubMed
      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          This paper discusses recommendations concerning the use of prior probabilities that underlie recent, but in no way novel, proposals of presenting scientific evidence in terms of posterior probabilities, in the context sometimes referred to as the 'full Bayes' approach'. A chief issue of this procedure is a proposal that--given the unavailability of case-specific circumstantial information--scientists should consider the prior probabilities of the propositions under which scientific evidence is evaluated as equal. The discussion presented here draws the reader's attention to the fact that the philosophical foundations of such a recommendation (in particular, attempted justifications through the Principle of Maximum Entropy (PME)) are far more controversial than what is actually admitted by the advocates for their use in the theory and practice of forensic science. Invoking only basic assumptions and the mathematical rules of probability calculus, the authors of this paper propose an argument that shows that there can be other more feasible and defensible strategies for eliciting reasonable prior probabilities. It is solely demanded that the reasoner is willing to make up his mind seriously on certain standard issues of fairly general criminal cases, such as evidential relevance or the probability of a suspect's guilt. However, because these issues intimately pertain to the responsibility of the trier of the fact, it is argued here that scientists' attempts to define appropriate prior probabilities should continue to be considered as untenable for the need.

          Related collections

          Author and article information

          Journal
          17267153
          10.1016/j.forsciint.2006.12.008

          Chemistry
          Bayes Theorem,Expert Testimony,Forensic Sciences,methods,Humans,Likelihood Functions,Models, Theoretical,Probability Theory,Shoes

          Comments

          Comment on this article