We implement causal mediation analysis using the methods proposed by Hong (2010) and Hong, Deutsch & Hill (2015) <doi:10.3102/1076998615583902>. It allows the estimation and hypothesis testing of causal mediation effects through ratio of mediator probability weights (RMPW). This strategy conveniently relaxes the assumption of no treatment-by-mediator interaction while greatly simplifying the outcome model specification without invoking strong distributional assumptions. We also implement a sensitivity analysis by extending the RMPW method to assess potential bias in the presence of omitted pretreatment or posttreatment covariates. The sensitivity analysis strategy was proposed by Hong, Qin, and Yang (2018) <doi:10.3102/1076998617749561>.
| Version: | 0.0.4 |
| Depends: | R (≥ 3.3.0), gtools, MASS |
| Published: | 2018-07-18 |
| Author: | Xu Qin, Guanglei Hong, and Fan Yang |
| Maintainer: | Xu Qin <xuqin at uchicago.edu> |
| License: | GPL-2 |
| NeedsCompilation: | no |
| CRAN checks: | rmpw results |
| Reference manual: | rmpw.pdf |
| Package source: | rmpw_0.0.4.tar.gz |
| Windows binaries: | r-devel: rmpw_0.0.4.zip, r-release: rmpw_0.0.4.zip, r-oldrel: rmpw_0.0.4.zip |
| macOS binaries: | r-release: rmpw_0.0.4.tgz, r-oldrel: rmpw_0.0.4.tgz |
| Old sources: | rmpw archive |
Please use the canonical form https://CRAN.R-project.org/package=rmpw to link to this page.