Base rates and the decision making model in clinical neuropsychology

Issues involving base rates and decision weighting techniques in neuropsychology are discussed. In both clinical and research work of neuropsychologists, the effect of asymmetrical base rate or unequal group size in research may have significant influence on the accuracy with which certain classifications are made. The general rule, given approximately unbiased error rates, is that the classification of greater incidence will be predicted more accurately than the low frequency classification. This disparity in predictive validity increases as asymmetry of base rate increases. Rather than despair about the decrease in overall predictive validity, the differential diagnostic accuracy can be utilized to estimate the probability of correctness of the obtained clinical decision.