The base rate fallacy controversy

This chapter focuses on the base rate fallacy controversy. The importance of considering base rates before making causal attributions is one that is instilled in all humans in the course of training in experimental methodology. But base rates play an important role in other inferential formats too, especially in Bayesian ones. There is evidence aplenty that in those contexts, they are largely ignored in favor of the diagnostic information at hand. This is the phenomenon known as the “base rate fallacy.” The probability of uncertain outcomes is often judged by the extent to which they represent their source or generating process. The features by which this similarity or representativeness is assessed are not necessarily those that figure in the normative derivation of the requested probability. Kahneman and Tversky derived the prediction that if an event is to be judged vis a vis several alternative possible sources or several alternative possible outcomes, these will be ranked by the similarity between them and the event and the ranking will not be affected by how likely each source or outcome is initially.