Transaction Search Form: please type in any of the fields below.
Date: November 25, 2024 Mon
Time: 8:24 pm
Time: 8:24 pm
Results for crime statistics (west virginia)
2 results foundAuthor: LaValle, Christina R. Title: Improving State Capacity for Crime Reporting: An Exploratory Analysis of Data Quality and Imputation Methods Using NIBRS Data Summary: Crime reporting in the United States originates from two major sources of data, the Uniform Crime Reports (UCR) and the National Incident-Based Reporting System (NIBRS). The incident-based reporting (IBR) structure of NIBRS is an enhancement to the traditional summary reporting of UCR used to track crime in the U.S. While the law enforcement community initiated IBR to address the expanding complexity of crime, reporting crime using NIBRS, like UCR, is voluntary and susceptible to issues of data quality, missing data, and noncompliance. Data collected using UCR and NIBRS are used for research and to document the status of crime at the national, state, and county levels. Data quality regarding accuracy and completeness are critical to reliable results and information. Further, reporting data "as is" without considering data quality and estimating for missing values may not be the most accurate depiction of the process and can result in criticism. As funding and resources lessen, coupled with the multitude of data fields involved with IBR, issues of data quality and missingness are areas of concern for analysts and researchers. In assessing data quality and handling missing data, appropriate and effective methods for resolving issues are necessary. Currently, elaborate methods established by the Federal Bureau of Investigation (FBI) are used to evaluate UCR data quality in terms of outlier detection (see Akiyama & Propheter, 2005). The FBI also imputes, or estimates, missing UCR data using methods that were developed in 1958 (Maltz, 1999). These methods are not timely, accurate, or easy for state programs to administer since they often use data from regions involving multiple states rather than the individual state. Moreover, methods have not yet been applied to aggregate crime count totals using NIBRS data and often reports using NIBRS data are criticized for being incomplete or non-representative. Details: Charleston, WV: West Virginia Division of Justice and Community Services, 2013. 37p. Source: Internet Resource: Accessed August 25, 2014 at: http://www.djcs.wv.gov/sac/documents/imputereport_rev4.pdf Year: 2013 Country: United States URL: http://www.djcs.wv.gov/sac/documents/imputereport_rev4.pdf Shelf Number: 129924 Keywords: Crime MeasurementCrime ReportingCrime Statistics (West Virginia)National Incident-Based Reporting SystemUniversity Crime Reports |
Author: LaValle, Christina R. Title: Testing the Validity of Demonstrated Imputation Methods on Longitudinal NIBRS Data Summary: The Uniform Crime Reporting (UCR) Program and the National Incident-Based Reporting System (NIBRS) are the two major sources of crime data in the United States. The UCR is a summary reporting system while NIBRS is an incident-based reporting system which was established to modernize crime reporting. The data collected by NIBRS is much more detailed. Given that law enforcement agencies across the nation voluntarily submit data to the Federal Bureau of Investigation (FBI) using either UCR or NIBRS, the presence of irregular reporting, missing data, and noncompliance are likely to compromise data quality. For many states, crime data collected using UCR or NIBRS are used to generate state and local crime reports and statistics. These data are most often reported "as is" and are thereby assumed correct. Since victimization data are typically not collected at the state or local levels to corroborate crime reports, there is an increased need for crime data to be as reliable as possible. Given the voluntary nature and inherent limitations of crime data collection systems, however, these data come with the caveat of being incomplete, or dubbed non-representative. Previous research on state incident-based reporting (IBR) data revealed issues with completeness, resulting from partial and non- reporting agencies, and accuracy, due to irregular reporting (LaValle, Haas, Turley, & Nolan, 2013). The previous work found that imputation, particularly alternative imputation methods developed by the West Virginia Statistical Analysis Center (WVSAC), can be used to reliably estimate for missing data. In conclusion of applying imputation methods to IBR data, the study also revealed that reporting data "as is" may not be the most accurate representation of IBR data. Additional studies have been conducted on national UCR data which found similar concerns with data quality, particularly issues related to missing data (see Maltz, Roberts, & Stasny, 2006; Targonski, 2011). Tools to detect and adjust for issues that are known to exist in crime data can improve data that are used as a basis for information and research. This research seeks to test and validate data quality techniques and imputation methods which will provide evidence that reliable and stable estimates of crime data can be attained with consistency over time. The study examines the performance of alternative imputation methods in comparison to FBI methods and provides a framework for the use of techniques on state-level IBR data. We apply and simultaneously test partial and non-reporting imputation methods using longitudinal data with the goal of improving the accuracy of state NIBRS data, especially when used for state and county trend analyses over time/ Details: Charleston, WV: West Virginia Division of Justice and Community Services, 2014. 30p. Source: Internet Resource: Accessed October 1, 2014 at: http://www.djcs.wv.gov/SAC/Documents/WV_Impute2ReportJan2014_Final.pdf Year: 2014 Country: United States URL: http://www.djcs.wv.gov/SAC/Documents/WV_Impute2ReportJan2014_Final.pdf Shelf Number: 133518 Keywords: Crime MeasurementCrime ReportingCrime Statistics (West Virginia)National Incident-Based Reporting SystemUniform Crime Reports |