R284

Variance Estimation and Selecting  Estimands for Causal Effect Queries

Anna K. Raichev, Rina Dechter, Jin Tian, and Alexander Ihler


Abstract

This paper investigates the statistical efficiency of algebraic expressions (``estimands'') used to answer causal queries.

We first examine structural rules, developing a partial dominance relationship for front-door estimands that extends recent results on back-door estimands.  

In many models, however, structural rules alone may be insufficient to select an estimand. 

For such cases, we propose empirical techniques for estimating and comparing the variance of different estimands

using both the graph and observational data: 1) a bootstrap-based method, and 2) a computationally simpler

yet practically effective method for discrete models which estimates the Fisher information matrices.

We illustrate both methods' effectiveness on a variety of causal diagrams and estimand forms.


[pdf]