Assessing the Quality of NIBRS Data for Generating Valid and Reliable Estimates of Crime by Place

• In Event: Improving Police-Community Relations through More Meaningful Data on Public Safety and Community Wellness

Abstract

The use of administrative data for statistical purposes is often hampered by insufficient documentation about how those data were collected and what the quality of the data are, at both the case- and reporting-unit levels. For the NIBRS data collection, the FBI conducts an extensive series of data quality checks that will reject a case if certain errors are found. Despite the checks built into the system, little is known about the underlying quality and variability of the NIBRS data within and across units. The limited research conducted to date reveals several anomalies in data values for which most consumers of the data do not account. Using NIBRS data from 2012-2014, this presentation will focus on a study of the quality of the NIBRS data including the distribution of data elements, the consistency of data coding, measurement errors in the data, and comparability of data over time. The impact of failing to adjust analyses to address these issues will also be discussed. The results of this project will provide guidance to practitioners and others on how to appropriately utilize these data and produce a better understanding of the nature of crime known to law enforcement in local communities.

Authors

• Alexia Cooper, Bureau of Justice Statistics, U.S. Department of Justice

• Kimberly Martin, Bureau of Justice Statistics, U.S. Department of Justice