That was my thought on reading the headline - it's somewhat skewed because you'd expect a similarly small percentage of the population to be responsible for _al

That was my thought on reading the headline - it's somewhat skewed because you'd expect a similarly small percentage of the population to be responsible for _all_ violent crime convictions (as most people aren't violent offenders; the study notes 3.9% had at least one violent conviction).