Web-based Software to Estimate Power and Causal Effects of Various Regression Discontinuity Designs

The purpose of this study is to develop user-friendly, web-based software for the analysis of regression discontinuity designs (RDDs). These designs are becoming increasingly popular for estimating causal effects in non-randomized studies. Their popularity in education research, however, has been slower in developing, perhaps due to the lack of user-friendly software for conducting the different analyses to which RDDs lend themselves.

The research team will develop power analysis software for sharp RDD, fuzzy RDD, RDD with multiple assignment variables, and regression kink designs. The software will also be able to conduct the analyses for estimating the causal effect in those types of designs. Researchers will use a Monte Carlo simulation study to test the software under a variety of modeling conditions, and once the software is ready, the researchers will create a demonstration of it using the Carolina Abecedarian Project and the Carolina Approach to Responsive Education, 1972–1992 (ICPSR 4091) study. Researchers will produce RDD software, which will be available online and presented at workshops and various conferences. The research team will also produce peer-reviewed articles.