`causalsandwich` is used to estimate causal effects with empirical sandwich variance estimator.

This means that the estimators have nice asymptotic properties when models are correctly specified.

Methods currently implemented

as of version 0.0.1:

Estimators of the average treament effect
IPTW
G-formula
Doubly Robust estimator

Why use `causalsandwich`? Easy implementation with valid inference!

This package is inteded to be a "one-stop-shop" with emphasis on ease of use.
The estimators here are consistent and asymptotically normal following usual regularity assumptions.

What's `geex`?

geex is a package designed for easy implementation of estimating equations. The causalsandwich package is powered by geex

What are estimating equations?

See Stefanski and Boos (2002)

How to install

The package can be installed from Github:

devtools::install_github("BarkleyBG/causalsandwich")