PSor: Semiparametric Principal Stratification Analysis Beyond
Monotonicity
Estimates principal causal effects under principal stratification
using a margin-free, conditional odds ratio sensitivity parameter. This
framework unifies the monotonicity assumption and the counterfactual
intermediate independence assumption, allowing for robust analysis when
monotonicity may not hold. Computes point estimates, standard errors, and
confidence intervals for conditionally doubly robust and debiased machine
learning estimators. The methodological details are described in Tong,
Kahan, Harhay, and Li (2025) <doi:10.48550/arXiv.2501.17514>.
| Version: |
0.1.0 |
| Imports: |
stats, SuperLearner, caret, dplyr, geex, magrittr, numDeriv |
| Suggests: |
testthat (≥ 3.0.0), knitr, rmarkdown |
| Published: |
2026-04-24 |
| DOI: |
10.32614/CRAN.package.PSor (may not be active yet) |
| Author: |
Jiaqi Tong  [aut,
cre],
Brennan Kahan [ctb],
Michael O. Harhay [ctb],
Fan Li [ctb] |
| Maintainer: |
Jiaqi Tong <jiaqi.tong at yale.edu> |
| BugReports: |
https://github.com/deckardt98/PSor/issues |
| License: |
MIT + file LICENSE |
| URL: |
https://github.com/deckardt98/PSor |
| NeedsCompilation: |
no |
| Language: |
en-US |
| Materials: |
README, NEWS |
| CRAN checks: |
PSor results |
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=PSor
to link to this page.