Assessing the sensitivity of methods for estimating principal causal effects

Research output: Contribution to journalArticlepeer-review

Abstract

The framework of principal stratification provides a way to think about treatment effects conditional on post-randomization variables, such as level of compliance. In particular, the complier average causal effect (CACE) - the effect of the treatment for those individuals who would comply with their treatment assignment under either treatment condition - is often of substantive interest. However, estimation of the CACE is not always straightforward, with a variety of estimation procedures and underlying assumptions, but little advice to help researchers select between methods. In this article, we discuss and examine two methods that rely on very different assumptions to estimate the CACE: a maximum likelihood ('joint') method that assumes the 'exclusion restriction,' (ER) and a propensity score-based method that relies on 'principal ignorability.' We detail the assumptions underlying each approach, and assess each methods' sensitivity to both its own assumptions and those of the other method using both simulated data and a motivating example. We find that the ER-based joint approach appears somewhat less sensitive to its assumptions, and that the performance of both methods is significantly improved when there are strong predictors of compliance. Interestingly, we also find that each method performs particularly well when the assumptions of the other approach are violated. These results highlight the importance of carefully selecting an estimation procedure whose assumptions are likely to be satisfied in practice and of having strong predictors of principal stratum membership.

Original languageEnglish (US)
Pages (from-to)657-674
Number of pages18
JournalStatistical Methods in Medical Research
Volume24
Issue number6
DOIs
StatePublished - Dec 1 2015

Keywords

  • complier average causal effect
  • intermediate outcomes
  • non-compliance
  • principal stratification
  • propensity scores

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability
  • Health Information Management

Fingerprint

Dive into the research topics of 'Assessing the sensitivity of methods for estimating principal causal effects'. Together they form a unique fingerprint.

Cite this