The 1,000 young people we describe in this book are members of an unselected birth cohort that represents a wide range of social origins. Since their births in 1972–3 they have been members of an ongoing longitudinal study called the Dunedin Multidisciplinary Health and Development Study. Their problem behaviour has been assessed repeatedly during their lives, as shown in table 2.1. Although in this book we examine data from all of the assessments, from age 3 to age 21, many of the analyses focus especially on the developmental period when participation in antisocial behaviour peaks, between ages 13 and 21. Thus, the data presented here describe the behaviour of the age cohort that contributed an extremely large proportion to their nation's crime rate between 1985 and 1995.

The Dunedin Multidisciplinary Health and Development Study is ideally designed for three types of research: (1) prediction studies of the childhood correlates of later health and behaviour outcomes, (2) developmental studies of continuity and change in health and behaviour, and (3) epidemiological studies of the prevalence and incidence of health problems and behaviour problems, and associations among problem types. This book presents all three types of analysis.

Description of the research sample and Study design Sample

The history of the Study and its design features have been described in detail in a book prepared by the team of investigators (Silva and Stanton, 1996).