Causal model
- X denote exposure, Y denote outcome, W denote measured covariates, Z denote unmeasured covariates
- Want to estimate the effect of different exposures on the outcome, which is quantified using the backdoor formula if W and Z close all backdoor paths.
- Above integrates over all measured and unmeasured covariates.
- Averages the true outcome distribution over the conditional distribution of the measured and unmeasured covariates.