Certification Tools: External Validity Checks

How better to know if NIBRS data are correct than to compare individual NIBRS incident data with the same incidents submitted separately to the state agency? This type of validity check is not only desirable but also necessary in Connecticut. Connecticut is committed to having a single NIBRS incident report replace all separate crime reports. This includes not only the UCR reports, but also Family Violence Offense Reports and Bias Crime Reports. Therefore, it is essential that all family violence and bias crimes be reported as Group A crimes in the NIBRS system. Therefore, comparing these reports with the NIBRS versions of the same incidents provides an evaluation of the presence and completeness of the NIBRS incidents.

Every certification contains this assessment. A copy of the details of Family Violence Offense Reports and Bias Crimes for a select time frame is printed out and the data are compared with the NIBRS incidents. Incident numbers provide the key for these comparisons.

Findings

The firs question is whether the incidents are Group A crimes, and then whether the data are comparable. Since the same officer usually completes the reports, discrepancies often point out systemic errors in coding offenses, injuries, weapons, or involved parties.

Download Connecticut's External Validity Review Checklist (.doc)