Propensity score - introduction and theorem

preview_player
Показать описание
This video provides an introduction to the concept of a 'propensity score' for an individual, and also states the propensity score theorem, which is a corollary of the conditional independence assumption.

Рекомендации по теме
Комментарии
Автор

I was coming very close to failing an econometrics class at university. Stumbling across your videos was the catalyst I needed to start to understand early lectures has enabled me to get a first in the first exam. There's a lot to do yet, but I now honestly believe I have a good chance at a first in econometrics, completely due to your help.

angusbruce-gardyne
Автор

Ben, amazing! You are able to explain literally any concept in Econometrics in such a simple way! I kept following your videos and whenever i feel lost between Google and textbooks, your videos get me the Heureka! Love it!

baconandphil
Автор

I am taking a 4th year econometrics class right now. My class is very fast paced, your videos really help me in understanding the underlying reasonings! Thank you!

feifeiyang
Автор

Your clear approach to explaining difficult concepts is amazing!

thomashaug
Автор

Thank you so much. I'm a big fan of your vids because they're super useful for trying to understand all the research stuffs that are completely unintelligible when trying to read from journal articles.

buckhum
Автор

An excellent introduction! Many thanks for these videos.

ujxtmap
Автор

Thank you so much! you are explaining it so simple and brilliant <3

stinefurst
Автор

How can you estimate a causal effect when this entire method is based on the assumption that there is no causal effect (the conditional independence assumption seems to imply that treatment Y and J are conditionally independent given X, which implies that X is causing both J and Y and J is not causing Y). What am I missing here?

marcosrodriguez
Автор

What samples do we use to form the logistic regression for the propensity? The exact samples of treatment group and non-treatment group that we want to match and compare?

AlexDodoChen
Автор

So....logistic regression to compress high dimensionality and sparsity/patchy data.

klam
join shbcf.ru