Causality: Instrumental Variables

The twelfth (and final??) video in a series on causality. This video covers instrumental variables, a way of using some sort of random variation in the world to give you something that looks a lot like an experiment! When it works. See more information about the class at nickchk.com/econ305.html.

[@tomz6327]

Thank you for explaining, and goodluck to anyone who is here to study for their upcoming exam!

[@Tbxy1]

I learned more from this 10 min video than from a 2-hour class.

[@fhoces]

Really enjoyed the videos. Strongly recommend for people that wants to learn more of the causal diagrams framework.

[@NickHuntingtonKlein]

Thank you!

[@gongxiaoli07]

Hi, Nick. If you have time, could you help me with a question: Is the ignorability assumption the same as the conditional independence assumption? Thank you.

[@NickHuntingtonKlein]

Sorry, never saw this for some reason. Ignorability as a concept is a bit more general than just conditional independence, but in application conditional ignorability and conditional independence, as applied to treatment assignment are basically the same.

[@learn_with_smaran]

What do you mean by back doors? does it mean confounding variables?

[@NickHuntingtonKlein]

Causal pathways from X to Y that aren't just arrows pointing away from X and towards Y. See my videos on causal diagrams.