NCJ Number
92344
Date Published
1983
Length
18 pages
Annotation
Local data sources generally list more cases of homicide than do FBI data, presumably because cases originally thought to be accidents, suicides, or 'undetermineds' are eventually classified as homicides at the local level but are not reported to the FBI.
Abstract
Study data came from FBI records and from police and medical examiner records in seven cities: Philadelphia, St. Louis, Chicago, Memphis, Dallas, Newark, and a California city which required anonymity. The analysis focused on the variables thought to have the highest levels of reliability: the total number of homicides; the age, sex, and race of the victim; and the weapon used to kill. The victim-offender relationship was also analyzed. The police and medical examiner records agreed in about 95 percent of the cases regarding the classification as homicide. Medical examiner departments tended to classify more cases as homicides than did police departments. The records also tended to agree on race, sex, victim's age, and weapon used. The greatest discrepancies among these concerned age. However, only 50 percent to 60 percent agreement existed regarding victim-offender relationships. The records tended to agree on family relationships, but police department sources had fewer listings of strangers or unknowns than did medical examiners. The areas of agreement and discrepancies between local and national records tended to parallel the patterns found in the comparisons of the two sets of local records. Researchers should pay closer attention to the data bases they use in the study of homicide and select the data base most appropriate to the variables they are studying. For age, total number of homicides, and victim-offender relationships, differences in results may partially reflect differences in the data base chosen. Data tables, notes and 11 references are provided.