Identifying High Risk Children

• In Event: Assessing Interventions across the Life Course

Abstract

Group-based trajectory modeling (GBTM) is designed to identify groups of individuals that follow a similar evolution in one or more measures over time. This paper applies a recent methodological advance that uses GBTM and its outputs to predict individual-level probabilities of an outcome of interest and estimate confidence intervals around these predictions, both of which can be updated in real time. The method is demonstrated with data from a large longitudinal Quebec-based study of individuals first assessed at age 7 in 1986-87. Based on various measures of psychological function (e.g., inattention) measured from age 7 to 11 we predict probabilities of their failing to complete high school and having a criminal record by age 22 at each age from 7 to 11. Use of the method for identifying high risk children for preventive interventions designed to avert these negative outcomes are discussed.

Authors

• Richard Tremblay, University of Montreal

• Daniel Nagin, Carnegie Mellon University

• Sylvana Cote, University of Montreal