Walls et al., 2011. Timescale Dependent Longitudinal Designs


In this chapter, we consider alternative temporal sampling designs within the interrupted time series family and explore other potential designs based on decisions about sufficient or optimal time scales capturing particular processes. We refer to this framework as time-scale-dependent longitudinal design (TDLD). A key objective is to provide a basis for researchers to develop studies of process that effectively utilize multiple time scales and intensive measurements. The chapter is organized in five sections: central issues, conceptual principles and statistical assumptions, developmental applications, an illustration, and future directions.