NCJ Number
206507
Journal
Homicide Studies Volume: 8 Issue: 3 Dated: August 2004 Pages: 214-254
Date Published
August 2004
Length
41 pages
Annotation
This study analyzed missing data problems in the Supplementary Homicide Reports (SHR) when imputing offender and relationship characteristics.
Abstract
The Supplementary Homicide Reports provide detailed, incident level data on nearly all murders and nonnegligent manslaughters in the United States. These reports include information on the month and year of an offense, on the reporting agency and its residential population, county and MSA codes, geographic division, population group, on the age, race, and sex of victims and offenders, and on the victim-offender relationship, weapon use, and circumstances of the crime. The record layout and variable definitions in the SHR data have remained almost unchanged since revision in 1976. The SHR form does not solicit certain key pieces of information and contains too little specificity in circumstance codes. The SHR is voluntary on the part of law enforcement agencies across the country therefore, agencies may not submit supplementary homicide data, and there is no easy way to track and capture these missing reports. The SHR also fails to provide full coverage of States. The study found that revising the weighting scheme currently used in the SHR data files held at the National Criminal Justice Data Archive (NCJDA) based on an improved set of covariates would increase effectiveness. Not only did eliminating year and State as classification factors avoid the problem of low frequency cells and thus outlier weight values, but the inclusion of additional covariates such as urbanness, region, and circumstances better accounted for deficiencies in known-offender data. In the future, and through NIBRS, efforts to classify homicide types should ultimately help the imputation process. The SHR limits the range of covariates to three basic victim characteristics, characteristics of the reporting agency, as well as limited and often questionable information about circumstances. Greater specificity in homicide data and additional efforts to understand the missingness in the homicide reports should enhance the ability to characterize and track patterns in homicide victimization and offending. Tables, references