Overlap between victimization and offending research in criminology: A weight activated main path analysis and topic network

• In Event: Policy Panel: Enhancing Translational Criminology with Natural Language Processing and Data Science

Abstract

Despite focusing on those most directly affected by the harms of crime, the subject of victimization is often treated as a distinct area of research within the study of crime. Indeed, the study of victimization is largely siloed from the study of offending in criminology. Such fragmentation of scholarship is especially surprising considering that victims and offenders are frequently one and the same. This phenomenon – dubbed the “victim offender overlap” – is one of the most well-documented and reliable findings in criminology, with decades of research confirming the substantial overlap between victimization and offending across a range of offending behaviors and contexts. To examine the extent that victimization research is integrated into the last 20 years of criminological research, the current study applies machine learning and network science methods to publication abstract and metadata collected from dimensions.ai. We use weight activated main path analysis (WAMPA) to identify the core of criminology and examine the distance between the main path of criminology and the main path of victimization. We then construct a topic network identifying the overlap between victimization research and other areas of criminology. Findings reveal a notable citation gap between victimization research and the main path of criminology. We present a topic network that introduces potential “zipper topics” to bolster the intersection of offending and victimization scholarship. Policy implications are discussed.

Authors

• Lindsay Leban, University of Alabama at Birmingham

• Thomas Bryan Smith, University of Mississippi

• Wesley G. Jennings, University of Mississippi