Return to search

Police Use of Force Databases: Sources of Bias in Lethal Force Data Collection

Understanding police use of lethal force requires the collection of reliable data. Due to bias present in police-use-of-lethal-force databases, researchers typically triangulate using multiple data sources to compensate for this bias; however, triangulation is restricted when the bias present in each database is unknown. This study investigates three government-funded and three independent police-use-of-lethal-force databases to identify methodological sources of bias present in the major U.S. data-collection systems. Bias was coded based on nine categories, including misclassification bias, broad conceptualization, narrow conceptualization, overlap bias, coverage bias, voluntary response bias, observer bias, gatekeeping bias, and self-report response bias. Findings suggest that all six databases had at least three different types of methodological bias present. Generally, public, government-sponsored databases exhibit bias through data self-reporting by law enforcement and varying victim race determination methods. Private databases reveal bias through media-based reporting and the triangulation of data from multiple sources, which is further complicated by lack of transparency in the databases' design and administrative procedures. All six databases have a unique position to the State, which should also inform researcher data selection. I argue that selecting data sources that complement each other based on these identified biases will produce a more complete image of police-use-of-lethal-force and enhance finding accuracy in future research. / Master of Science / Understanding incidents where a civilian dies due to the actions of police officers requires the collection of reliable data. Due to bias—flaws in the data collection methods or data presentation—which lead to varying results when using different databases, researchers typically use multiple data sources to make up for these flaws; however, this method is restricted when the bias present in each database is unknown. This study investigates three government-funded and three independent police-use-of-lethal-force databases to identify sources of bias present in the major U.S. data-collection systems. Findings suggest that all six databases had at least three different types of flaws present. Generally, public, government-sponsored databases exhibit bias through police self-reporting lethal force, where an officer's department reports the officer's actions and there is no individual or group outside of the police reporting these incidents. Additionally, there is a flaw in how police record the race of a victim, who dies through police use of lethal force; Varying procedures in how race is recorded, whether recorded based on the officer's opinion or where a victim self-reports their own race prior to death on a government data system such as the Department of Motor Vehicles, also impacts the race data included in public databases. Private databases reveal bias through collecting incident data from news reports and using data from multiple sources such as law enforcement reports, medical examiner reports, and media reports simultaneously; this is further complicated by lack of transparency in the databases' design and administrative procedures, where there are no documents detailing the steps databases take in collecting and presenting data. All six databases have a unique position to the U.S. Government, where some are funded by the Government and where some are motivated by recent high profile police killings, which should impact researcher data selection. Ideally, the databases used should hold multiple perspectives or positions to the Government to provide an more complete image of lethal force. I argue that selecting data sources that complement each other based on these identified biases will produce a more complete image of police-use-of-lethal-force and enhance finding accuracy in future research.

Identiferoai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/103615
Date28 May 2021
CreatorsWalkup, Christian Andrew
ContributorsSociology, Reichelmann, Ashley Veronica, Rocha Beardall, Theresa, Hawdon, James E.
PublisherVirginia Tech
Source SetsVirginia Tech Theses and Dissertation
Detected LanguageEnglish
TypeThesis
FormatETD, application/pdf, application/pdf
RightsIn Copyright, http://rightsstatements.org/vocab/InC/1.0/

Page generated in 0.0016 seconds