Mean IQ in the group was 155, although their method of determining this was a bit iffy (in some cases they used SAT/GRE as an IQ proxy). MOre than 70% held a college degree, so these are not necc. marginalized gifted drop-outs.
However, they seem to have whacked up the data a little bit by oversampling men who were incarcerated. They address this, and it does take away some of their conclusions. One interesting fact that remains:
. For concealed weapons, bomb making, and aggregated offenses, the female high-IQ sample reported significantly greater prevalence rates than the control group. Interestingly, prevalence rates of the female high-IQ sample appear to more resemble the male high-IQ and control groups’ rates than those of the female control group.
Basically, I think we are left with:
Even after omitting incarcerated subjects from the analysis and separating males from females, high-IQ males reported significantly higher prevalence rates than controls for homicide and attempted suicide; high-IQ females reported significantly higher prevalence rates than controls for concealed weapons, bomb making, and aggregate violent crime.
But it's a methodologically flawed study, to say the least. Nonetheless, it does suggest that crime without punishment is certainly going on among the HG.