likelihood that noncompelled reporters will approach sentinels for child abuse, relative to elder abuse. Thus, sentinels for child abuse have access to greater conduits of information than their older adult counterparts.
Overall, the unresolved issue of mandatory reporting of elder abuse, the relative infancy of elder abuse public education, and the limited conduits of information on elder abuse cases flowing to potential sentinels may severely limit the application of this form of child abuse assessment to elder mistreatment in that the method may lack sensitivity. This lack of sensitivity will be particularly problematic for the population of non-cognitively impaired, relatively independent mistreated older adults who wish to avoid formal service agency involvement in their abuse situations.
There are several sources of victim statistics describing rates of violent crime in this country (e.g., National Crime Victimization Survey [NCVS], Federal Bureau of Investigation [FBI] Uniform Crime Reports [UCR], FBI National Incident Based Reporting System [NIBRS]). Official police or government estimates of assaultive violence are typically lower than those obtained by social scientists conducting epidemiological research. These differences are largely attributable to methodological variance across surveys (e.g., use of gateway versus behaviorally specific preliminary screening questions, or aggregation of official police reports versus population surveys, see discussion of this below). This variance is informative: failure to use direct, behavioral questions leads to failed case identification.
The FBI’s UCR is a frequently cited index of violent crime that has been reported to police. The UCR is a case-based report, in which the worst FBI index crime (murder, rape, robbery, aggravated assault, burglary, larceny, motor vehicle theft, arson) reported by an individual is the only one that is recorded for that individual. However, since many crimes are not reported to police, and because many individuals are multiply victimized, UCR results are somewhat misleading.
The Bureau of Justice Statistics overcomes this weakness in its annual NCVS of approximately 80,000 to 100,000 adults aged 12 years and older from approximately 45,000 households. Randomly contacted U.S. citizens are asked about both reported and unreported victimization experiences. In 1992, older adults (age 65 years and older) comprised 14 percent of survey respondents (Bachman, 1992). According to the NCVS, adults over age 50 were the least likely to be physically or sexually assaulted, with an annual violent crime rate of 12.5 per 1,000. However, once assaulted, older adults were more than twice as likely to be seriously injured and require