Koehler, J. J. (1993). The Normative Status of Base Rates at Trial. In Castellan, N. J. (Ed.). Individual and Group Decision Making. p. 137-149. Hillsdale, NJ: Erlbaum.

ABSTRACT

This paper was largely concerned with legal criticisms about the probative merit of base rate probability evidence. The traditional criticisms were found to be wrong or unconvincing. The claim that probabilities are diagnostically relevant to a large series of cases but irrelevant to the individual cases that compose the series is logically indefensible. Similarly, the argument that would treat overtly probabilistic evidence as diagnostically inferior to evidence that is only implicitly probabilistic cannot be sustained; probative value is not negatively related to numerical explicitness.

Although probability evidence can and should be defended against these criticisms, special problems associated with base rate evidence were identified. Most worrisome are the difficulties associated with identifying appropriate reference classes, and treating target cases as if they were sampled at random from these reference classes. These problems are serious and deserve more attention than they have received. On the other hand, it was argued that these problems do not necessarily justify disregarding base rates in favor of other types of evidence as some have suggested. Instead, the probative value of base rates that are not obtained under ideal circumstances should be treated as an empirical matter. Hopefully, future studies will investigate the important prescriptive issues related to identifying the conditions under which attentiveness to these suboptimal base rates will and will not improve judgmental accuracy in the courtroom and elsewhere.