11.2 - Causal Discovery from Interventions in Two-Variable Setting

In this part of the Introduction to Causal Inference course, we introduce causal discovery from interventions in the simple two-variable setting. Please post questions in the KZbin comments section.
Introduction to Causal Inference Course Website: causalcourse.com
Course Lectures Playlist: • 11.2 - Causal Discover...

[@blueonification]

Thanks for these videos on intervenional cd! I have actually waited for them and as expected, they are pure gold. May I ask, why you investigate only the two unidirected causal relations and causal independence, but not the bidirected causal relation (A⇔B)?

[@BradyNealCausalInference]

I'm glad you share my enthusiasm! I was also particularly looking forward to this week. I'm actually not aware of literature on the bidirected causal relation (not saying there isn't any). In fact, I don't have intuition for how they can be the case. The common class of examples I can think of on this (where A causes B *and* B causes A) is where the dimension of time is collapsed. In other words, A causes B at some point in time and B causes A at some point in time. And the way I commonly hear this resolved is "unravel the time dimension," which might give you a graph like A_0 -> B_0 -> A_1 -> B_1 -> A_2 -> B_2 -> ...