Regression Discontinuity Design (RDD or RD)

Regression Discontinuity Design (RDD or RD) is a quasi-experimental research method used to measure the causal effects of some variable by assigning individuals, samples, or other units to a control and treatment group when crossing some preset threshold. This method is used in the analysis of randomized experiments with a fractional treatment, or when the experimenters are unable to jointly randomize and assign the treatment. RDD studies rely on the assumption that the outcome of the treatment group and the control group will be distinct in the region immediately prior to the threshold, but that the two groups become nearly indistinguishable on the other side of the boundary. By contrasting the outcomes in the two groups, RDD allows for the study of the effect of a treatment near the cutoff.