The assumptions underlying statistical analysis are rarely fully met. Pragmatists face the challenge of knowing when evidence is strong enough to justify action, and that’s where sensitivity analysis helps by expressing the robustness of inferences relative to potential biases. For instance, sensitivity indices can quantify how much of the observed effect would need to be bias-driven to alter conclusions. We are building on these ideas by developing the methods and associated resources listed on this page.
Questions? Issues? Suggestions? Reach out through the KonFound-It! Google Group.
Start KonFounding
Try out KonFound-It! to calculate sensitivity analyses through an interactive web app.
Explore Benchmarks
Explore sensitivity analyses calculated for the What Works Clearinghouse through an interactive web app.
Get the latest version of the konfound package:
Code to run:
install.packages("devtools")
devtools::install_github(
"konfound-project/konfound"
)
Code to run:
ssc install konfound, replace
ssc install indeplist, replace
ssc install moss, replace
ssc install matsort, replace