Linking commercial burglaries by modus operandi: tests using regression and ROC analysis

Area of Science:

• Criminology
• Forensic Science
• Statistical Analysis

Background:

• Police departments frequently use modus operandi (MO) to connect crimes to a single offender.
• Empirical validation of MO's effectiveness in linking crimes is crucial for forensic analysis.
• Understanding behavioral consistencies in criminal activity aids in offender profiling and crime linkage.

Purpose of the Study:

• To statistically evaluate the reliability of modus operandi (MO) in linking commercial burglaries to common offenders.
• To identify specific behavioral features that effectively differentiate between linked and unlinked crime pairs.
• To assess the predictive accuracy of these behavioral features in comparative case analysis.

Main Methods:

• Analysis of data from 86 solved commercial burglaries involving 43 distinct offenders.
• Application of logistic regression analysis to identify significant behavioral discriminators between crime pairs.
• Utilization of Receiver Operating Characteristic (ROC) analysis to quantify the predictive accuracy of identified MO features.

Main Results:

• Certain behavioral features demonstrated high predictive accuracy in distinguishing linked from unlinked crime pairs.
• The distance between burglary locations emerged as a particularly strong indicator of offender consistency.
• The study identified significant consistencies in commercial burglary behavior.

Conclusions:

• The findings support the validity of using modus operandi (MO) for crime linkage in commercial burglaries.
• Behavioral consistencies, such as spatial proximity of offenses, offer theoretical insights into criminal activity.
• The identified features can form the basis for a diagnostic tool to aid comparative case analysis in law enforcement.