Knowledge, probability, and credibility

This paper concerns study of the credibility of human sources of evidence and its relation to the inferential value of testimony they provide. From a certain view of 'knowledge' in epistemology comes the suggestion that credibility assessment can be construed as a cascaded inference in which attributes of human source credibility are identified. Scholarship from evidence law in jurisprudence suggests an evidential basis for credibility assessment in terms of these attributes. Applying Bayes' rule to this cascaded inference offers a way of expressing and combining credibility‐related beliefs in the process of assessing the inferential value of evidence.