Statistical inference and graphical procedures for RD designs using local polynomial and partitioning regression methods.

Manipulation testing using local polynomial density methods.

Finite-sample inference for RD designs using local randomization and related methods.

Power and sample size calculations for RD designs using robust bias-corrected local polynomial inference.

Estimation, inference, RD Plots, and extrapolation with multiple cutoffs and multiple scores RD designs.

We analyze two student outcomes – student GPA and average ACT test score for 1000 12th graders at the given school district. Specifically, we want to understand, among all the information we know about these students – demographics (gender, ethnicity, special education status, age, household income, and performance index), AP course taking, GPA …