Illustrating the stability of conduct problems, two different developmental pathways have been found to predict different outcomes into adolescence, with child-onset having greater negative consequences. A representative cohort of 457 boys were studied from the time they were three years old until they were 18 to determine if there was a distinction between child-onset and adolescent-onset conduct problems (Moffitt et al., 1996). The sample was from the Dunedin Multidisciplinary Health and Development Study in Dunedin, New Zealand. Over 90% of the sample was of European ancestry (less than 7% were Maori or Polynesian). The children were assessed every two years from baseline until study completion. It was found that those who were on a life-course-persistent path (i.e., the individual had conduct problems from childhood into adolescence) and those who were on an adolescent-limited path did not differ in their symptoms in early adolescence. However, those in the life-course-persistence group had more deviant characteristics than those on the adolescent-limited path in childhood. The life-course-persistence group also had greater instances of convictions for violent crime, more instances of school leaving, and poorer bonds with their families. Similar results were found in a separate study in New Zealand, the Christchurch Health and Development Study (Fergusson, Lynskey, & Horwood, 1996). In order to examine the developmental continuity of conduct problems from childhood to adolescence, 901 children were first assessed when they were between 7 and 9 years old and followed-up when they were either 15 or 16 years old. The researchers found that children with early indicators of conduct problems were 16 times more likely to have conduct disorders as adolescents, compared to those who did not show evidence of conduct problems at the initial assessment.