causalfrag
Runs, classifies, interprets and reports sensitivity analyses for unmeasured confounding across the partial R-squared robustness value approach (Cinelli and Hazlett, 2020, <doi:10.1111/rssb.12348>), E-values (VanderWeele and Ding, 2017, <doi:10.7326/M16-2607>), and the impact threshold for a confounding variable and robustness of inference to replacement (Frank, 2000, <doi:10.1177/0049124100029002001>; Frank, Maroulis, Duong and Kelcey, 2013, <doi:10.3102/0162373713493129>). An ordinary least squares crosswalk reports the robustness values, impact threshold and replacement percentage computed from the focal t statistic and residual degrees of freedom, makes explicit that their agreement is largely fixed by that shared input, and flags the boundary band in which they disagree. Template-based plain-language reports are included, with optional integration with the 'confoundvis' package for plots.
- Impact of a Confounding Variable on a Regression Coefficient
- Making Sense of Sensitivity: Extending Omitted Variable Bias
- Sensitivity Analysis in Observational Research: Introducing the E-Value
- What Would It Take to Change an Inference? Using Rubin’s Causal Model to Interpret the Robustness of Causal Inferences
This page was built for software: causalfrag
