Using temporal network methods to reveal the idiographic nature of development.

Abstract

Averages dominate developmental science: There are representative groups, mean trajectories, and generalizations to typical children. Nearly all parents and teachers, however, eagerly proclaim that few youth are average; each child, adolescent, and young adult is unique. Indeed, individual youth are the focus of many eminent developmental theories, yet there is a shocking paucity of developmental methods-including study designs and analysis techniques-that truly afford individual-level inferences. Thus, the goal of this chapter is to explicate the advantages of an idiographic approach to developmental science, that is, an approach that provides insight into individual youth, often by studying within-person variation in intensive longitudinal data, such as densely coded observations, repeated daily or momentary assessments, and functional neuroimages. In three domains across development, the chapter illustrates the benefits of an idiographic approach by comparing empirical conclusions offered by traditional mean-based analysis techniques versus techniques that leverage the temporal and individualized nature of intensive longitudinal data. The chapter then concentrates on group iterative multiple model estimation (GIMME), which is an analysis technique that uses intensive longitudinal data to create youth-specific temporal networks, detailing how brain regions or behaviors are directionally related across time. The promise of GIMME is exemplified by applications to three different domains across development. The chapter closes by encouraging future idiographic developmental science to consider how research questions, study designs, and data analyses can be formed, implemented, and conducted in ways that optimize inferences about individual-not average-youth.