The largest and most confounding variable is the extent to which traumatization preceded, and perhaps even caused, the act of killing. Posttrauma symptoms, especially rage and numbing, can themselves be causal in acts of violence. The converse then follows: of those who commit violence, a disproportionate number may well have suffered prior traumatization compared to the general public.
For criminal homicides, Pollock (1999) looks at this directly by ascertaining a group that did not experience previous traumatization and therefore does not have this as a confounding variable. For Vietnam veterans, an article looking at premilitary factors found no personality or traumatic predecessors for those who engaged in “abusive violence” (Hiley-Young et al., 1995). It was the level of combat that most predicted engagement in such behavior. Especially when killing is socially sanctioned or even socially obligatory, there will be a set of people for whom prior traumatization is not an issue.
However, because there will also be those for which prior traumatization is a confounding issue, leaving this group out will lead to only partial understanding. Some studies on PTSD have tried to eliminate the confounding variable of alcohol abuse by removing those suffering such abuse from the experiments. While this does give more details about what is attributable only to the PTSD, it gains an understanding only of those without alcoholism. Since substance abuse problems are commonly associated with PTSD, a major portion of the population has simply remained unstudied. Similarly, the role of prior trauma-