Michael Johns: Propensity Score Matching: A Non-experimental Approach to Causal... | PyData NYC 2019

[@danielechavezs7502]

00:53 Outline 01:30 Propensity Score Matching in a Nutshell 03:25 The Selection Problem 05:52 Key Steps 07:18 Goals

[@rembautimes8808]

Good lecture by an expert in the field and the slides were well designed as well. Great talk

[@GuorongLi-re7kt]

One question, if we get overfitting propensity scores, then the overlap we want will be very small. It looks like conflict arguments here.

[@surayuthpintawong8332]

May I ask what the conflict is? If you are overfitting, in the extreme case, each person is assigned only one propensity score. Thus, it's nearly impossible to find anyone with a similar score.

[@Akerfeldtfan]

Helpful lecture by data science Tucker Carlson!

[@MMUnubi]

😀😀