Causality: Closing Back Doors

[@SigneMacholm]

This explanation was super helpful to understand how to approach front door and back door adjustment. THANK YOU!

[@jjman897]

Thank you for your videos, very underrated. After reading the textbook and watching videos from my teacher I was completely lost until I got to your videos on the basics of causality

[@nataliemcglynn153]

This is an amazing series. You explain very complex and confusing concepts, so clearly and it's finally clicking for me. How do you not have more views!!!

[@NickHuntingtonKlein]

Thank you! And good question

[@kunalkohli5392]

Thanks for this amazing series. I recently started working on causal inference and was lost after reading the book of why. This series really made everything so simple to understand

[@user-wr4yl7tx3w]

Just found this channel. Great content.

[@tomaszkozakiewicz4451]

Wow! Nobody explained it that clearly to me!

[@andrewhordern2283]

Great, clear video in an excellent series.

[@econometria2637]

Excellent explanation, thanks!

[@garbour456]

Amazing. Crystal clear. Thanks a bunch!

[@crazyabtmusic26]

thank you so much..this video was very helpful ..finally understood the backdoor thing..

[@qiguosun129]

Thanks for the clear explanation!😀

[@erice5025]

Excellent explanation!

[@erice5025]

If I just read the book by peral, it would take my life

[@qinghuafeng1705]

Really helpful, thank you very much! Now I know why controlling for drugs is a bad control. But could you let me know why you call confounding factor as "back door"? Why it is "back"? Thank you.

[@NickHuntingtonKlein]

Thanks! The back door terminology comes from the fact that it's an alternate way you can get from cause to effect. On a causal diagram, you can follow arrows pointing from cause to effect (for example cause -> outcome) - those are front doors. But there are two ways to get out of your house - the front door or a back door! Back doors are alternate way to get from cause to effect (for example cause outcome)

[@koenoosthoek7882]

Thanks a lot for your videos! They help me out a lot. I just wondered: at 5:35, you mention that you could go from wine, to health, to life. How is this possible, when the wine arrow is not pointed towards health ( H --> W )? This is only possible if the wine arrow were to point towards health right? ( W --> H)?

[@koenoosthoek7882]

My question also applies to the other back door paths

[@NickHuntingtonKlein]

@@koenoosthoek7882 a causal path is any way on the graph that you can get between two variables, and gives one reason those variables might be related. So it doesn't particularly care which direction the arrows point (unless there's a collider, that's another video). The paths you're talking about - where all the arrows point away from treatment - would be ways that treatment causes outcome, but we also want the ones that describe how treatment and outcome are spuriously related

[@koenoosthoek7882]

@@NickHuntingtonKlein Thank you for your quick response, I'll just ignore the direction of the arrows when looking for back door paths :-)

[@haseebali512]

Why does controlling for health alone close the last two backdoors @7.45? Would you not be required to control for both health and income to be able to close these backdoors?

[@NickHuntingtonKlein]

Controlling for any one variable on the path closes the whole path (as long as it's not a collider - that's another video). You can think of it this way - if you have a path like, for example, x

[@user-wr4yl7tx3w]

Why everything from the perspective of wine?

[@NickHuntingtonKlein]

It's just an example

[@user-wr4yl7tx3w]

What if arrows go both ways?

[@NickHuntingtonKlein]

See this section of my book theeffectbook.net/ch-DrawingCausalDiagrams.html#avoiding-cycles