Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Bayesian Inference for Fuzzy Regression Discontinuity Designs
Alessandra Mattei, Fan Li, Fabrizia Mealli

Last modified: 2013-06-16

Abstract


Regression discontinuity (RD) designs are usually interpreted as local randomized experiments. We formalize this concept proposing a probabilistic formulation of the assignment mechanism. We develop a Bayesian approach to draw inferences for the causal effects around the threshold, focussing on a fuzzy RD design where the receipt of the treatment is based on eligibility criteria and a voluntary application status. The method is applied to evaluate the effects of Italian university grants on student dropout.

Full Text: PDF