Understanding Bias and Item Missing Data in NIBRS

• In Event: Overcoming Measurement Challenges

Abstract

Data produced by the FBI’s National Incident-Based Crime Reporting System (NIBRS) offers researchers, policy analysts and practitioners exceptional opportunities for understanding crime. Many assume, however, the data are biased due to lack of national coverage, over representation of smaller population areas, and incomplete reporting of item data. This effort attempts to measure the bias in NIBRS data and establish how that bias is changing over time as law enforcement agencies increasingly employ NIBRS reporting. Using our relational database implementation of NIBRS, which contains over 74.6 million crime incidents, we examine NIBRS data over a sixteen year period and compare NIBRS crime breakdowns and crime rates for violent and property crimes with national UCR Summary Program results. We also examine how missing item data has changed over this period. Finally, we describe several tools that we have developed to allow researchers to explore this large, complex collection of incident crime data.

Authors

• Douglas Salane, The Graduate Center, John Jay College / CUNY

• Eman Abdu, The Graduate Center, John Jay College / CUNY

• Peter Shenkin, The Graduate Center, John Jay College / CUNY