Dear All,
Thanks for a great discussion today! There indeed was a mistake on slide #72 where it should have said that: "Implicit in most discussions of potential outcomes is that the outcome for a given unit under a specific treatment does not depend on the treatment given to any other unit, i.e., the independent outcomes assumption." The revised slides are now posted online. We will also account for this in the homework grading.
If you want to learn more about stability and other assumptions necessary for the counterfactual model, I posted an article by Cole and Frangakis 2009 in the section Recommended Readings for week 1.
Have a great weekend,
Lydia