NCJ Number
235465
Journal
Corrections Compendium Volume: 35 Issue: 3 Dated: Fall 2010 Pages: 1-6,12,13
Date Published
2010
Length
8 pages
Annotation
This study examined the compatibility and quality of research that uses single and multiple data sources from the core accounting files provided by the U.S. Bureau of Justice Statistics (BJS): the Survey of Inmates in State Correctional Facilities, the National Corrections Reporting Program (NCRP), and the National Adult Correction Census.
Abstract
The analysis of the admissions and release data between 1984 and 2000 showed evidence of comprehensive prison records. The inter-survey analysis showed the content and coverage quality of the prison data for several indexing variables. The NCRP provides individual-level data with background and descriptive characteristics of interest for the study of crime. For the prison population, these variables can be used with relative confidence in many cases. The quality of parole population data, however, requires further testing. The parole analysis produced evidence of coverage errors. Although by definition the parole data provided all records for each State reporting in a given year, the analyses in this study found evidence of under-reporting by as much as 20 percent in the years studied. Future data analysis should continue to verify the improvement of intercensal and intersurvey data quality for the overall population and by sex. Researchers should be aware of the degree of consistency of the available data and be attentive to the plausibility of the assumption that the survey data and the NCRP data reflect one another varies across indexing variables. Given the significant implications of analyses with these data, every effort should be made to know and understand the limitations of combining the sources; continually assess the quality and compatibility of the sources; and seek ways to improve the current quality and ease of merging these different data sources. 4 tables, 2 figures, and 51 references