ABSTRACT
This chapter presents the concepts of coherence and correspondence as criteria for appraising the quality of probability judgments. It focuses on advanced approaches, which are critical to understand human judgment, and coherence and correspondence biases often arise when correspondence-based mechanisms are assessed in terms of coherence-based standards. For simplicity of exposition, probability is illustrated with exact figures, but the generality of the argument should be clear. Some see coherence criteria as a fundamental shortcoming of the coherentist approach, and argue that correspondence criteria are the appropriate means for appraising probability judgments. Others argue that the coherence constraint is sufficient. Bayes' rule provides a normative rule for updating one's probability judgments in the light of new evidence. To illustrate the use of Bayesian reasoning in a legal context, the case of DNA evidence is used. The Bayesian approach is illustrated with examples from medical and legal decision making, and extended to more complex contexts involving multiple variables.
