Neighborhood Human Mobility Clustering and Its Impact on Crime: Advancing Spatial Analysis by Taking a Closer Look at the Characteristics of Ambient Population

• In Event: Communities and Place: Crime and Fear of Crime: Exploring Emerging Methodologies

Abstract

This study examines the link between neighborhood human mobility patterns, representing ambient populations, and their effects on property, violent, and drug-related crimes. It addresses the limitations of traditional crime analysis models focused on static residential populations by investigating the dynamic nature of ambient populations across census block groups (CBGs) in Arlington, TX.

Using geocoded data from Safegraph human mobility tracking in 2019, the research applies routine activity theory with an innovative approach, integrating geographic factors like CBG counts of device travel origins, median distance, dwell time, device-to-stop rates, and daytime ambient population proportions. The study presents separate analytical models for different crime types using spatial multivariate clustering analysis, showing how attribute fields on neighborhood mobility patterns concentrate within Arlington's CBGs.

Additionally, ANCOVA is used to evaluate how various mobility clusters affect different crime types while controlling for traditional social disorganization variables. This research provides a more thorough understanding of how transient populations' complex mobility patterns influence crimes, improving the precision of spatial crime prediction.

Authors

• Olga Semukhina, Tarleton State University

• Junghwan Bae, Tarleton State University

• Stan Korotchenko, Tarleton State University

• Chris Copeland, Tarleton State University