Predicting the Primary Residence of Serial Sexual Offenders: Another Look at a Predictive Algorithm

This study analyzes a series of crimes committed by sexual offenders and a series committed by non-sexual offenders. Subjects chosen for the study were seven sexual offenders (Test Group) and twelve offenders with non-sexual crimes (Control Group). Test data were selected from the Minnesota Predatory Offender Registry. Control group data were randomly selected from data provided with the crime analysis software used in this study (CrimeStat III). Historical offense data from all offenders were analyzed using the Journey-to-Crime (JTC) function of CrimeStat III to predict the most likely location of residence based upon the location of offenses. The distance predicted was compared with the actual location of the residence of each offender. Data collected for each group was compared both visually and statistically to determine if there was a significant difference between the two groups.