Causal inference Reflections / Post-Mortem

[nickeubank]

• The reading reflections need to have guided questions. Students are struggling to know what to focus on (and probably need more accountability)
• Students are struggling to see how things fit together. See these notes for overview lecture. More of that sooner I think.
• Emphasize that, in a sense, the ATE v. ATT distinction is kinda about "generalizability" / external validity.
• Add pre-post from Morgan and Winship before diff-in-diff.
• Offer Morgan and Winship / Imbens and Rubin potential outcomes readings as supplements?
• Maybe drop gender/race examples? #38 Doesn't quite comport with "manipulation" emphasis.
• Maybe 1 Kohani Tang and Xu reading in initial experiments presentation?
• More on weighting and why we put matching data into regression.