search

Kuan-Liu-Lab / causens

An R package for causal sensitivity analysis methods
https://kuan-liu-lab.github.io/causens/
MIT License
1 stars 0 forks source link

Binary outcome for the first version of the sensitivity analysis #9

Closed Kuan-Liu closed 5 days ago

Kuan-Liu commented 9 months ago

Binary outcome has better statistical property, the senstivity function derived using binary outcome is bounded by -1 and 1:

$$ \text{sens} = E(Y^a \mid A=1, X) - E(Y^a \mid A=0, X) .$$

The bounds allow us to custermize prior for the bias parameters for the Bayesian approach and the specification of the sensitivity function of the frequentist Li's approach.

larryshamalama commented 9 months ago

Okay sounds good

larryshamalama commented 7 months ago

I looked into continuous outcomes, this issue addressed specifically binary outcome variables