Data Science Research Methods: Causality and selection

In this presentation, I talk about the potential outcomes framework, how one can define causal effects, the fundamental problem of causal inference, selection, and how one can overcome it using experiments. Then, I talk about how one can analyze outcomes from experiments, what can go wrong, how one can provide supportive evidence for randomisation, and how one can use instrumental variables estimation to account for imperfect compliance.

[@tayoukachukwu3105]

Very useful and insightful video. Learn't some important points. Thanks for sharing. Hope to see more

[@wouter2462]

nice

[@littledog1dog759]

Take a shot each time he says um, uh, uuuuh or sth of the sort. you'd be dead after this