Carl C. Bell, M.D.
Community Mental Health Council, Inc.
Department of Psychiatry, College of Medicine,
University of Illinois at Chicago
The relationships among contagion, group marginalization, and resilience form a complex issue that does not lend itself to quantitative methodology, but rather is best studied using qualitative methods. Thus, having a historical perspective is an important attribute to understand appropriately the phenomenon of violence as it relates to contagion, group marginalization, and resilience or protective factors. Furthermore, in order to have a coherent discussion about violence, we must first understand which type of violence we are focusing on, as violence is a very complex and multi-determined phenomenon. In addition, we must understand the science.
The Need for Good Science
To understand this complex problem, we must understand the need for good science. Unfortunately, there is a fundamental scientific problem with understanding violence, whether it is directed toward others or self-directed. The reality is that these phenomena, while being the third or leading cause of death for some population groups, such as teens or young black males, respectively, are actually rare events. The reality is that suicide rates tend to be 11 suicides/100,000 (IOM, 2002) and homicide rates are about 9/100,000 (Douglas and Bell, 2011). Even if you focus on non-Hispanic black people who have rates of homicide of around 33/100,000 or gun homicides with rates of 58/100,000, these are low base rates and developing statistical power to differentiate between an experimental intervention and control is very difficult. Accordingly, the 2002 Institute of Medicine report Reducing Suicide: A National Imperative noted that to prove a suicide prevention intervention is evidence-based, a study would need 5 to 10 population studies with 100,000 persons per study to get enough statistical power to show that either a suicide or homicide prevention study works (IOM, 2002). Because the homicide rates are actually lower than the suicide rates, despite many scientific claims to the contrary, apparently one would need an equally large population to prove a homicide prevention intervention is evidence-based, and neither of these two studies has been done.