Judgment and Decision Making

PART III

Judgment

Bayesian inference is a general approach to making orderly inferences about uncertain events. It assumes that people have prior beliefs about any question – if only a state of ignorance. When they encounter new data, they interpret it in light of their general knowledge about such situations, leading to updated (or posterior) beliefs. Bayesian inference treats all beliefs as subjective, even if there is wide agreement about them (e.g., the chances of winning a fair roulette play). However, it insists that beliefs be consistent with one another and with gambles based on them, before calling them probabilities. As a result, subjective does not mean arbitrary. Bayesian inference provides the foundation for the normative analyses ...