Criminal evidence indexes for taxometric investigations of guilty-suspect base rates and classifications

Abstract Taxometric methods may help researchers estimate guilty-suspect base rates in the criminal justice system and develop procedures for classifying individual suspects. This program requires that criminal evidence be quantified for statistical analysis, but no appropriate indicator variables exist. We discuss the criteria for suitable taxometric indicators and outline characteristics that indexes of criminal evidence should meet. We derive quantitative indexes of four types of criminal evidence: eyewitness identifications, interrogations and confessions, forensic match evidence, and propensity evidence. We sketch how the proposed indexes could be validated within the taxometric program. Further, we illustrate how researchers can empirically probe critical assumptions underlying these indexes, quantify the sizes of departures from these idealizations, and leverage these estimates to modify and improve index calculations. The proposed indexes address an outstanding impediment to the realization of the taxometric program to study guilty-suspect base rates in the criminal justice system.