Back to Projects List
Improving Evaluation Methods for Targeted Educational Interventions
Principal Investigator: HyeonJin Yoon
Funding Agency: Office of Research and Economic Development—Layman Award
Award Date: Aug 1, 2020
End Date: Jul 31, 2021
Regression discontinuity design (RDD) has a great utility for evaluating cut-score based educational interventions because it allows educators and policymakers to target students most in need as well as make a robust causal inference for the intervention impact. However, a key limitation of the basic RDD is that its causal inference is only warranted to the area near a cutoff, which is seldom of research or policy interest.
The study examines the utility and statistical validity of a new evaluation method for targeted educational intervention for students at risk — RDD with covariate matching (RDD-CM).
The project aims to demonstrate the application, analysis and interpretation of the RDD-CM to identify causal effects beyond the treatment cutoff, and to examine the extent to which an RDD-CM yields unbiased and precise estimates comparable to those from a randomized controlled trial (RCT) design.
Investigation will be conducted using kindergarten math intervention data. Findings will inform education researchers and policymakers in designing and implementing practical and ecological evaluation plans that allow for robust causal inference and ethical delivery of interventions.
Additionally, the project will provide necessary empirical evidence to submit an external methodological research grant to National Science Foundation’s Faculty Early Career Development (CAREER) Program.