This awardee has received supplemental funding. This award detail page includes information about both the original award and supplemental awards.
Description of original award (Fiscal Year 2010, $9,313,707)
This solicitation seeks to design and conduct a major survey to accompany the National Crime Victimization Survey (NCVS). In response to an interest among our stakeholders for the production of sub-national estimates, this project is intended to lay a foundation for determining the most viable and cost-effective option for the development and implementation of a large-scale effort to generate sub-national crime victimization estimates.
The primary goal of this solicitation is to generate preliminary data at the MSA-level to test the feasibility of producing estimates of the annual incidence of victimization in these areas. It is anticipated that, ultimately, the sub-national companion data collection will strategically boost the core NCVS sample, and that the blending of the two data collections will provide the necessary statistical power to produce reliable estimates at the sub-national level. Another key goal of this research is to develop the means to create adjustments to the MSA-level estimates to account for the differences between the sub-national companion survey methodology and the core NCVS methodology. Given that BJS expects the survey response rates for this companion survey to be lower in owing to light of the lower-cost methodology and as such subject to a higher level of non-response bias, BJS plans to leverage the data from the core NCVS, which currently achieves nearly a 90 percent response rate through in-person interviewing techniques, to improve the non-response adjustments for the sub-national companion data. Ultimately, this will reduce the non-response bias in the blended survey data. The cost and viability of the sub-national companion survey methodologies in these areas will be compared to the estimates, metadata, and para-data of the core NCVS.
The objective is to blend the data from this survey with the core NCVS to produce reliable and valid sub-national estimates of the incidence of victimization at a reasonable cost. The key objectives for this study are to:
1. Develop a sub-national companion data collection to boost the core NCVS and evaluate the companion collection for reliability, validity, cost-effectiveness, and sustainability.
2. Evaluate the feasibility and utility of blending the sub-national companion data into the core NCVS, despite the markedly different sources and levels of coverage, response, and measurement error.
3. Compare the trade-offs in response bias, cost, operational complexity, and estimation between the core NCVS and a lower-cost, sub-national component.
4. Develop sufficiently reliable data at the sub-national level to develop and evaluate model-based estimation procedures that could serve as an additional component to the survey estimates.
This project seeks to design and conduct a major survey to accompany the National Crime Victimization Survey (NCVS). In response to an interest among our stakeholders for the production of sub-national estimates, this project is intended to lay a foundation for determining the most viable and cost-effective option for the development and implementation of a large-scale effort to generate sub-national crime victimization estimates. The primary purpose of this project is to generate data at the MSA-level to test the feasibility of producing estimates of the annual incidence of victimization in these areas. The primary goals are to: 1.Use a relatively inexpensive design and methodology; 2. Create a survey design that could be administered by local jurisdictions or their vendors; and 3. Produce estimates of value to the local area that correlate with similar estimates available from the national NCVS, including trend estimates. The project will compare the trade-offs in response bias, cost, operational complexity, and estimation between the core NCVS and a lower-cost, sub-national component. To date, the project team has developed a mail survey design using address-based sampling, with a household informant reporting on victimization experienced by adults in the household. One form of this survey supports only person- and household-level estimates, while the other also supports incident-level estimates. A Pilot Test of this design was successfully fielded in Fall 2015, with sampled addresses from the 40 largest Core-Based Statistical Areas (CBSAs). A second wave is planned for Fall 2016. This award will allow sufficient sample to continue the evaluation of the design, including its ability to measure local victimization trends. Wave 2 of the Pilot Test will also include several methodological experiments. Two experiments continued from Wave 1 are testing the two questionnaire versions and inclusion of Spanish-language materials for addresses not associated with Hispanic surnames or areas with large Hispanic populations. Wave 2 will also test different levels of cash incentive ($2, $1, $0) and the use of FedEx versus USPS for the second follow-up questionnaire mailing.
Note: This project contains a research and/or development component, as defined in applicable law.