Other public health analysts, finding these traditional public health models somewhat confining, have shifted to a "behavioral science model" that orients them to "complex social prevention" (Rosenberg et al., 1986:1399-1400). The justification for adopting this perspective has been stated by Rosenberg et al. (1986:1400):
Public health as a discipline must progress beyond the traditional disease model, with its natural tendency to overemphasize the importance of purely medical problems and interventions. … A new terminology must be developed for prevention that allows for multiple levels of determination on the one hand, and, on the other, for interventions that are designed less to change individual behavior or the natural environment than to change social behavior and the social environment.
Although this approach is promising, it looks less like traditional epidemiology and more like sociological and criminological methods for understanding the causes of violence. Indeed, in this conception, the public health analysis of violence comes into almost perfect alignment with criminological and sociological views of the problem.
Although it is true that the criminal justice practitioner community tends to focus on the character and intention of offenders as the most important causal factor determining incidents of victimization, the criminal justice research community has long engaged in analyses of interpersonal violence, and criminal offending more generally, that are in the same spirit as epidemiological analyses carried out by the public health community. It is also true that, increasingly, criminal justice practitioners are joining them in this understanding of criminal violence and are basing their operational strategies on these conceptions.
For example, for several decades, sociologists and criminologists have done empirical work disclosing patterns in the incidence of criminal victimization and criminal offending (e.g., see Wolfgang, 1958a; Hindelang et al., 1978). Over the last decades, that work has been enormously advantaged by the routine use of victimization surveys (Dodge, 1981). Such work will also be advanced