Predicting Police Stops: Integrating Self Reports and Community Data in a Multilevel Model

Kirk Miller, Northern Illinois University
Matthew T. Zingraff, North Carolina State University

The poster presents results from multilevel modeling procedures that account for driver and community-level factors that influence the likelihood of being stopped by police. Disaggregating self-reported stops by local police versus the North Carolina State Highway Patrol yields patterns reflecting the differential organizational goals, connection to local structural conditions, and primacy of driver and driving characteristics of these two types of police organizations. Self report data from a telephone survey of 2920 licensed drivers in North Carolina conducted in 2000 is integrated with county level 2000 Census data to perform the hierarchical linear models.

(Return to Program Resources)

Updated 05/20/2006