The Interchangeability of Homicide Data Sources: A Spatial Analytical Perspective

The authors extend the work of Wiersema, Loftin and McDowall (2000) by assessing the interchangeability of Supplemental Homicide Report (SHR) and National Center for Health Statistics (NCHS) homicide rates in the context of spatial analysis. Conclusions drawn from spatial analyses may be affected by the choice of data source if the underrecording of homicide tends to cluster in geographic space more strongly in one data source than the other. Such a situation could alter indications of homicide rate clustering and substantive conclusions drawn from multivariate spatial regression models. Results indicate that although these data sources are interchangeable for urban counties, spatial effects and the effects of social structural factors vary by data source for rural counties.