An influence function projection approach exploits graph-implied conditional independences to improve the efficiency of semiparametric estimators for upper and lower bounds on average causal effects under sensitivity models for unmeasured confounding.
Journal of the Royal Statistical Society Series B: Statistical Methodology , volume=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Exploiting independence constraints for efficient estimation of bounds on causal effects in the presence of unmeasured confounding
An influence function projection approach exploits graph-implied conditional independences to improve the efficiency of semiparametric estimators for upper and lower bounds on average causal effects under sensitivity models for unmeasured confounding.