The web app "Regression Sensitivity Analysis: the Robustness Value and the Partial R2"
the functionalities of the R package sensemakr
available on the web.
This web app allows researchers to examine how sensitive their regression estimates are to the inclusion of unobserved confounders, as well as bounding the strength of unobserved confounders based on comparisons to observed covariates. To perform these analyses, the user only needs to provide summary statistics already found in standard regression tables. These results do not rely on any assumptions regarding the functional form of the treatment assignment mechanism nor on the distribution of the unobserved confounders, and hold for multiple confounders, possibly acting non-linearly. For details, please see Cinelli and Hazlett (2020)