This report describes an assessment that focused on state-level estimates produced within the expanded National Crime Victimization Survey; it details the external context as well as internal validation, including coverage, nonresponse, measurement, data processing, and sampling errors.
This report describes an assessment of the redesigned National Crime Victimization Survey (NCVS) sample design to (1) identify potential sources of error that may compromise the validity of state-level estimates (i.e., internal validation) and (2) evaluate how NCVS estimates compare with and augment estimates from the Federal Bureau of Investigation’s Uniform Crime Reporting (UCR) Program (i.e., external context). The internal validation assessed potential errors with coverage, nonresponse, measurement, data processing, and sampling. The external context focused on comparing NCVS victimization estimates for crimes reported to the police with analogous estimates from the UCR. No overarching issues were identified during this examination that would call into question the validity of state-level victimization estimates produced from the redesigned NCVS sample. A state-level NCVS analysis user’s guide will also be available to approved analysts in the Census Bureau restricted-use data environment with guidance on generating victimization estimates and standard errors.