function_ideas.md

In benjamin-ackerman/generalize: Generalizing Average Treatment Effects from RCTs to Target Populations

Function Ideas:

1: Trim target population to match covariate bounds of trial

2: Calculate trial participation probabilities using

  - GLM
  - Random Forests
  - Super Learner?
  - Lasso
  - others?

3: Compare distributions of trial participation probabilities using

  - Tipton generalizability index
  - Kolmogorov-Smirnov?
  - others?

4: Create plots (?) or at least covariate tables?

  - Covariate difference b/w trial and population pre-and-post weighting
  - Plot trial participation probabilities
  - Plot SATE and TATE, show difference between the two

5: Generalize results using

  - Weighting by the odds
  - BART
  - TMLE

benjamin-ackerman/generalize documentation built on Oct. 11, 2020, 3:58 a.m.