S3E27: Propensity Scores -- I Meant To Do That!
Quantitude
00:00
The Propensity Score, the Predicted Probability of Treatment
There are a few ways to adjest for the propensity score. Ther's like four canonical ways. One way which i don't recommend, is you actually treat the propensi score as a confounder and you adjust for it using ancova. So that's one reasona i don't like that is cazsette's just going back to the old use of ancova,. We're gong to leave that behind. Are you an ancovist? You sound like an anCovist. It makes me think back to harold hotel lane in the early 19 thirties with principal components analysis. He did not care about rescaling igan vectors
Transcript
Play full episode
