Skip to content

Code to reproduce the findings in "Bayesian Model Averaging in Causal Instrumental Variable Models"

Notifications You must be signed in to change notification settings

gregorsteiner/gIVBMA-Code

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This repository contains the code to reproduce all the findings in "Bayesian Model Averaging in Causal Instrumental Variable Models", arxiv:2504.13520.

The applications folder includes all the code to reproduce the empirical examples. In particular, Carstensen_Gundlach.jl, schooling.jl, and birthweight.jl contain the code for the geography or institutions, returns to schooling, and birthweight examples, respectively. The priors folder includes code to reproduce the plots illustrating the priors implied by the hyperprior on $\nu$. The simulations folder includes all simulation experiments included in the paper:

  • simulation_KO2010.jl includes the code for the setting with many weak instruments, and simulation_pln.jl includes the code for the design with a Poisson endogenous variable.
  • simulation_Kang2016.jl includes the code for the setting with some invalid instruments.
  • simulation_MultEnd.jl includes the code for the setting with multiple endogenous variables and correlated instruments.

The gIVBMA.jl package (https://github.com/gregorsteiner/gIVBMA.jl) is required to run the experiments and applications.

About

Code to reproduce the findings in "Bayesian Model Averaging in Causal Instrumental Variable Models"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published