Description of original award (Fiscal Year 2008, $226,256)
In 2007, BJS sponsored an expert panel study by The National Research Council of the National Academies of Science to review the National Crime Victimization Survey's (NCVS) methodology and provide recommendations for redesign options in conducting the NCVS. In its interim report, the panel stressed that BJS should carefully study potential changes in study design before incorporating them (Groves & Cork, 2008). BJS has initiated a methodological research program to support the redesign of the NCVS based on panel recommendations.
The NRC evaluation noted that costs of data collection have forced BJS to reduce sample size to the point that the survey can no longer measure year-to-year changes in criminal victimization with adequate precision.
The objective of this project is to develop sample designs which improve the efficiency of the NCVS sample.
The project will:
1. Evaluate what variables on the sampling frame of the NCVS could be used to reduce sample variance while maintaining a representative sample.
2. Determine what external variables could be used to reduce standard errors.
3. Recommend over-sampling strategies which could be implemented to improve estimate precision.