Publications & Resources
The Use of Piecewise Growth Models in Evaluations of Interventions
Michael Seltzer and Martin Svartberg
In studies of interventions (e.g., remedial reading interventions), interest often centers on student academic progress, or on changes in various attitudinal and affective measures, both during and after the intervention period. By enabling us to subdivide a time series into meaningful segments, and summarize important aspects of change in each segment, piecewise growth models provide a means of addressing key questions in intervention studies. In this report, we discuss the use of piecewise models in (1) examining whether rates of progress for individuals in an intervention study, on average, slow down, remain constant or speed up during the follow-up period; (2) assessing whether there is substantially more variability among individuals in their rates of change in the intervention period or in the follow-up period; (3) identifying conditions under which we see rapid rates of progress during the intervention period, and sustained progress during the follow-up period.
Seltzer, M., & Svartberg, M. (1998). The use of piecewise growth models in evaluations of interventions (CSE Report 477). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).