Bridged or Not Bridged

• Activity Number: 526
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Jieru Chen*+
• Companies: CDC
• Keywords: race ; violence ; reporting

Abstract

This study assesses the effect of race categories in the reporting of injury surveillance data. The analyses were based on data collected by the National Intimate Partner and Sexual Violence Surveillance System developed for measuring violence-related behaviors in 50 states and the District of Columbia. The statistical model for bridging, an algorithm developed at the National Center for Health Statistics, was applied to one year of the surveillance data. The model included demographic and contextual covariates, some at the county level and others at the person level, and was used to bridge multiple-race categories into single-race categories. Analysis results show that bridging has the most effect on the American Indian and Alaskan Native (AIAN) population, followed by Other Non-Hispanic and Asian Non-Hispanic populations, respectively. Among persons reported physical violence, sexual violence, or stalking victimization, bridging is seen to have an large impact on the AIAN and Asian Non-Hispanic populations. In addition to assessing impact of bridging on prevalence, impacts of bridging on the associations of race/ethnicity with experience of violence history were also assessed.