The Bureau of Justice Statistics (BJS) and the FBI are collaborating with RTI International to develop a new set of statistical processes required to estimate the incidents and characteristics of crime and arrest in the U.S. These new procedures are based on data from the FBI’s National Incident-Based Reporting System (NIBRS). Currently, about half of all law enforcement agencies do not contribute their crime data to NIBRS. It is critical that these processes address the coverage gap and other data quality considerations in the reported data.
Theoretical data quality assessment frameworks were utilized first to identify potential quality issues and possible remedy strategies. Statistical adjustment procedures were then developed and implemented into the estimation process to mitigate the identified issues. Last, final adjusted data were used to produce estimates and a variance estimation procedure was created. This presentation will discuss these procedures, as well as provide an overview of the initial estimates generated from these processes. Presenters will also compare the newly-developed NIBRS estimation procedures to those historically applied to summary-based crime data.
|