Drew Bailey is a developmental psychologist by training. His research focuses on the processes underlying the longitudinal stability of individual differences in children’s mathematical achievement and on the medium- and long-term effects of educational interventions. His current work attempts to use psychological theories and methods to build models to improve the accuracy of predictions about the medium- and long-term effects of educational interventions. The goal is to understand why educational interventions often produce effects that diminish after the end of the intervention and to identify combinations of interventions and populations likely to generate the most persistent effects.
My plans for the fellowship period
Random assignment evaluations of academic skill-building interventions often show promising initial impacts that all but disappear within the next few years. Surprisingly, this pattern has been found in the domain of early mathematics, despite theory and empirical work suggesting early math interventions might be expected to generate strong and persistent effects on later math skills. Theoretically, early math skills are clearly foundational to children’s later math performance and learning. Empirically, strong correlations between children’s early and much later math achievement have proven to be robust to a variety of statistical controls across several non-experimental datasets.
The discrepancy between the optimistic theoretical and correlational predictions on the one hand and disappointing patterns of intervention effects on the other raises two crucial questions for promoting children’s math trajectories: i) how do we identify the most important and durable early math skills to target with interventions; and ii) which populations are most likely to enjoy long-run benefits from early math interventions?
Over the course of the fellowship period, I will attempt to answer these questions by leveraging causally-informative designs from experimental studies of early math interventions to validate methods for producing causally-informative estimates of the longer-run impacts of early math skills using non-experimental data.
For each of the experimental datasets, I will estimate patterns of impacts on available domains of mathematical achievement. Second, using only the control groups from the experimental studies, I will estimate a variety of correlational models to reproduce the patterns of impacts found in the experimental data. I will then apply the most promising of these methods to several longitudinal, non-experimental datasets to identify the early skill targets and child characteristics most likely to generate persistent positive impacts on children’s math achievement.
How will my work change children’s and youth’s lives?
A validated method for forecasting long-term effects of early math interventions using non-experimental data should prove enormously valuable for designing early interventions that promote children’s academic development. Because the general problem of forecasting the long-term effects of interventions is specific neither to the domain of early mathematics nor to children in any particular country, the methods developed in this project could be applied to other academic domains as well.