A residential program I visit has a system for monitoring severe behavior problems, such as aggression toward other persons, running away, or self-injury: Staff fill out a form describing the behavior then a data entry person puts the info into a database.
The graph above is simply a count of incidents involving aggression to persons. The x-axis is the age of the client at the time the incident happened. These raw data are from 2008.
The above graph confirms my clinical impression: with respect to aggression, the teens and young adults are champs (I should mention that about one-third of the clients at this facility are under and two-thirds are over 18 years of age).
If I were a clever doctor, I'd try to take new patients of about 21 years. I could straighten these kids out with friggin' Fred Flintstone's Chewable Multivitamins. People would think I was a genius.
See why we need control groups for clinical research involving kids?