Propensity score weighting procedures for causal inference with clustered data

In the three articles that make up this dissertation, we studied the behavior of treatment effect estimators that take into account the multilevel structure of the data. The estimators we studied accomplish this by modeling the relationship between treatment assignment and confounders at both the unit and cluster levels. More specifically, they estimate a function of the sample called the propensity score, which is a summary of the confounding information. The literature on propensity score procedures offers a variety of ways to use the propensity score to control for confounding, but here we focus on propensity score weighting estimators, which weight the sample using weights constructed from the propensity score prior to computing a treatment effect.

Rechte

Nutzung und Vervielfältigung:


CC BY-ND 4.0

Bitte beachten Sie, dass einzelne Bestandteile der Publikation anderweitigen Lizenz- bzw. urheberrechtlichen Bedingungen unterliegen können.

Zitieren

Zitierform:
Zitierform konnte nicht geladen werden.