Strengthening the regression discontinuity design using additional design elements: A within-study comparison

This paper compares the regression discontinuity design (RDD) to the randomised clinical trial approach and proposes adding untreated comparisons to RDD to mitigate weaknesses in the method.

The study demonstrated that through the use of an untreated comparison in the basic regression discontinuity design, it is possible to overcome three of the main problems associated with it compared to randomized control trials (RCT):

  • Lower statistical power
  • Dependence on statistical modelling assumptions
  • The limiting effects of its treatment effect estimates due to the narrow subpopulation of cases immediately around the cutoff.

Sources

Wing, C. and Cook, T. D. (2013). Strengthening The Regression Discontinuity Design Using Additional Design Elements: A Within-Study Comparison, Journal of Policy Analysis and Management, 32(4):853-877. Retrieved from: https://onlinelibrary.wiley.com/doi/abs/10.1002/pam.21721

'Strengthening the regression discontinuity design using additional design elements: A within-study comparison' is referenced in: